Article Text


Causes and outcome in cough

Statistics from


1H. Sumner, 1A. Kelsall, 2A. L. Lazaar, 1U. Kolsum, 1D. Singh, 1A. A. Woodcock, 1J. A. Smith. 1Respiratory Research Group, University of Manchester, Manchester, UK, 2Respiratory CEDD Discovery Medicine, GlaxoSmithKline, USA

Introduction Patients with chronic obstructive pulmonary disease (COPD) complain of cough and mucus hypersecretion. The mechanisms underlying cough are likely to be multifactorial and are poorly understood. We aimed to investigate the effect of smoking on objective cough frequency in both patients with COPD and healthy volunteers.

Methods We studied 68 patients with COPD from the ECLIPSE cohort (median age 66 years (interquartile range (IQR) 61–71), median forced expiratory volume in 1 s (FEV1) 59.5% predicted (IQR 44.3–75.3), 22 female, 23 current smokers), 12 healthy smokers (median age 60 years (IQR 57–60), median FEV1 91.6% predicted (IQR 43.7–71.0), 7 female) and 12 healthy non-smokers (median age 58 years (IQR 43–68), median FEV1 117.0% predicted (IQR 103.0–122.8), 8 female). All subjects underwent 24 h objective cough monitoring (Vitalojak, Vitalograph), spirometry and capsaicin cough reflex sensitivity testing.

Results Significant differences in 24 h cough rates were seen between subject groups (p<0.001). COPD current smokers had the highest cough rates (median 9.0 coughs/h (IQR 4.3–15.6)); COPD ex-smokers and healthy smokers had similar cough rates (median 4.9 coughs/h (IQR2.3–8.7) and 5.3 cough/h (IQR 1.2–8.3), p = 0.60) which were about half those of COPD current smokers (p = 0.018 and p = 0.03). Healthy volunteer cough rates (median 0.7 coughs/h (IQR 0.2–1.4)) were lowest. Objective cough rates were not influenced by age or gender. Cough reflex sensitivity was not significantly different between groups (logC5 p = 0.35, logC2 p = 0.36) but was weakly correlated with cough frequency (logC5 r = −0.36, p = 0.004 and logC2 r = −0.31, p = 0.001). Smoking history and FEV1 % predicted also correlated with cough frequency (r = 0.49, p<0.001 and r = −0.40, p<0.001). In a linear regression model 47.7% of the variance in cough frequency (across all subjects) could be explained by smoking history (p = 0.002), current smoking status (p = 0.002), FEV1 % predicted (p = 0.002) and cough reflex sensitivity (logC5 p = 0.016).

Conclusion Ambulatory objective monitoring provides novel insights into the relative effects of smoking and airflow obstruction on the symptom of cough.

On behalf of the ECLIPSE Investigators.

Funding: Funded by GlaxoSmithKline (CRT110639, SCO104960, NCT00292552).


1P. A. Marsden, 2B. Ibrahim, 1L. Yates, 2A. Woodcock, 2S. J. Fowler, 2J. A. Smith. 1University Hospital of South Manchester, Manchester, UK, 2University of Manchester, Manchester, UK

Rationale There is evidence to suggest that subjectively scored cough is related to asthma control. This study aimed to explore the relationships between objective cough rates, airways inflammation and disease control in asthma.

Methods 89 subjects with physician-diagnosed asthma were studied (mean age 57.4 (±12.0) years; 57.3% female; 7.9% smokers; 39.3% ex-/52.8% non-smokers; median asthma duration 29.0 years (10.0–68.0); mean forced expiratory volume in 1 s (FEV1) % predicted 86.4% (±22.1); median dose response ratio (DRR) to methacholine 24.5 (0.0–1970.1); median inhaled corticosteroid (ICS) dose 800.0 μg (0.0–4000.0); median sputum eosinophils 2.0% (0.0–26.0)). Subjects underwent 24-hour ambulatory cough monitoring with the VitaloJak cough monitor; coughs were manually counted and expressed as number of explosive cough sounds per hour (cs/h). In addition, subjects completed the Juniper Asthma Control Questionnaire (ACQ) and underwent sputum induction.

Results Cough recordings were obtained in 96.6% and sputum samples in 61.8% of subjects. Median cough rates were higher by day (3.7 cs/h (0.2–41.3)) than by night (0.5 cs/h (0.0–29.6)) (p<0.001) and median ACQ score was 1 (range 0–4.4). Objective cough rates correlated with ACQ scores by day (r = 0.28; p = 0.008), night (r = 0.24; p = 0.027) and over 24 h (r = 0.33; p = 0.003). However, there were no significant correlations between cough rates and either sputum eosinophils (%: r = 0.16; p = 0.26; cells ×106/g sputum: r = 0.14; p = 0.31) or neutrophils (%: r = −0.09; p = 0.50; cells ×106/g sputum: r = −0.06; p = 0.66). Similarly there were no significant correlations between ACQ scores and sputum eosinophils (%: r = 0.20; p = 0.15; cells ×106/g sputum: r = 0.26; p = 0.054) and neutrophils (%: r = 0.02; p = 0.90; cells ×106/g sputum: r = 0.07; p = 0.61). In a linear regression model, the percentage of eosinophils (p = 0.005), smoking pack year history (p = 0.005) and overall cough rate (p = 0.008) explained 38% of the variance in ACQ scores (p<0.001; adjusted R2). In a similar regression model, eosinophils ×106/g sputum (p<0.001), smoking pack year history (p = 0.005) and overall cough rate (p = 0.002) explained 47.9% of the variance in ACQ scores (p<0.001).

Conclusions Higher cough rates indicate poor asthma control. Sputum eosinophils, 24 h cough rate and smoking history independently predict asthma control, and number of eosinophils corrected for sputum weight explains more of the variance in ACQ scores than percentage eosinophils.


1N. Yousaf, 2S. S. Birring, 1I. D. Pavord. 1Glenfield Hospital, Leicester, UK, 2Kings College Hospital, London, UK

Background Up to 40% of patients seen in a cough clinic have unexplained chronic cough. The long-term outcome of these patients is uncertain.

Objective To determine the long-term outcome in patients diagnosed with unexplained chronic cough.

Methods We have performed a longitudinal study of symptoms, airway inflammation and spirometry in a cohort of patients with unexplained chronic cough diagnosed >7 years ago. Cough was assessed using a 100 mm visual analogue scale (VAS). At the final visit cough reflex sensitivity was assessed as the concentration of inhaled capsaicin at which the volunteer coughed two (C2) and five times (C5).

Results We identified 42 patients (32 females) with unexplained chronic cough who had been assessed at least twice over at least 7 years. The mean (SD) duration of cough was 11.5 (4.5) years at the time of their final assessment. Nine patients (21%) had organ-pecific autoimmune disease and 20 (48%) had a peripheral blood lymphopenia. Six (14%) patients had complete resolution of symptoms and 17 (40%) had a significant >15 mm improvement in their cough VAS during follow-up. Longitudinal spirometry data were available in 30 patients. The geometric mean rate of forced expiratory volume in 1 s (FEV1) decline was 63 ml/year and four (13%) patients developed a postbronchodilator FEV1/forced vital capacity (FVC) of <0.7. Log C2 and log C5 at the time of final assessment were significantly correlated with the log rate of FEV1 decline (−0.71, p = 0.014 and −0.70, p = 0.018). FEV1 decline was similar in patients with persistent cough and those whose cough improved. No other independent predictors of FEV1 decline were identified. There were no independent predictors of improvement in cough.

Conclusions Cough improves over time in the majority of patients with unexplained chronic cough. However, patients have an increased rate of decline in FEV1 which is associated with a heightened cough reflex.


S. Faruqi, P. Sedman, A. H. Morice. Castle Hill Hospital, Cottingham, UK

Introduction and Objectives Extraoesophageal reflux is a common cause of chronic cough. This is often refractory to medical therapy. Surgical treatment in the form of a Nissen fundoplication (NF) can be potentially curative. There is paucity of data regarding response to NF in patients with chronic cough.

Methods We retrospectively reviewed the case notes of patients from our Cough Clinic who had undergone NF over the past 6 years. Demographic details, duration of symptoms, presence of other symptoms, results of oesophageal studies, outcome and complications were recorded.

Results 49 patients (mean age 54 years, females 37) underwent NF from May 2003 to April 2009. The average duration of cough was 8 years. Gastrointestinal symptoms were present in the majority; however, they were mild in most and not an indication for NF by themselves. Seventeen patients were on “asthma” treatment as well. All patients had oesophageal pH and manometry studies performed prior to surgery. 35 patients had significant acid reflux and 2 had severe oesophageal dysmotility. 31 (63%) patients had a response to treatment, 21 good and 10 partial. 10 of 17 with asthma felt that the asthma had improved as well. Mild dysphagia or bloating was seen in 17 patients following surgery. Five patients needed repeat surgical intervention for modification of NF. One patient with severe chronic airways disease and occult cardiac disease developed aspiration pneumonia 8 weeks following surgery and died.

Conclusion Our response rate of 63% is consistent with that observed by others. Our cohort of patients had cough as the indication for surgery. These patients were otherwise refractory to multiple medical therapeutic trials. Surgical treatment in refractory disabling cough is a valid option. Surgical treatment does have significant associated complications and patients should be carefully selected. This is a controlled study and hence a placebo effect cannot be ruled out.

View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.