Prevalence and Causes of Cough in Chronic Dialysis Patients: A Comparison Between Hemodialysis and Peritoneal Dialysis Patients

Frank Min, Susan M. Tarlo, Joanne Bargman, Naveen Poonai, Robert Richardson, Dimitrios Oreopoulos

Peritoneal dialysis (PD) and hemodialysis (HD) are both common forms of dialysis for patients with end-stage renal disease. A few case reports have suggested that cough is associated with PD. From 1991 to 1998, 17 patients being treated with PD at the Toronto Western Hospital demonstrated persistent cough severe enough for referral to a respirologist. Causes of cough, often more than one cause per patient, included asthma, post-nasal drip, gastroesophageal reflux disease (GERD), chronic obstructive pulmonary disease, congestive heart failure, allergic rhinitis, pleural effusion, and respiratory infection. The aim of this cross-sectional study was to establish the prevalence of cough among PD patients, to determine if PD patients more commonly have a dry persistent cough than do HD patients, and, if the latter case is true, the possible reasons for it.

A detailed survey of 92 PD patients and 91 HD patients was conducted in 1998 and 1999 at the University Health Network. Survey questions inquired about patient respiratory symptoms since onset of dialysis. Charts were reviewed to obtain information on use of medications possibly relevant to cough.

In the PD and HD groups, 52% and 23% were females (p = 0.001), and the mean ages were 59.1 and 60.1 years, respectively. Angiotensin converting enzyme (ACE) inhibitors had been taken by 65% (PD) and 55% (HD) of patients, and beta-blocking medications by 43% (PD) and 51% (HD). Since initiation of dialysis—mean 2.7 years (PD) and 3.7 years (HD)—22% of PD patients reported persistent cough versus 7% of HD patients (p = 0.003). Although no significant association was seen between cough and self-reported heartburn in HD patients (p = 0.439), a significant association between cough and self-reported heartburn was seen in PD patients: 67% of PD patients with persistent cough reported heartburn versus 29% of those without cough (p = 0.008). The findings suggest that GERD and associated cough are more common in PD patients than in HD patients, perhaps owing to increases in intra-abdominal pressure from the peritoneal dialysate.

Key words

Cough, renal failure, gastroesophageal reflux disease


Peritoneal dialysis (PD) is a common form of out-of-hospital dialysis. A slow but ongoing expansion of its use has taken place in recent years, particularly thanks to the development of automated equipment that makes PD more flexible, comfortable, and efficient. It might be expected that patients receiving PD would be at increased risk of chronic cough.

To treat chronic hypertension, many dialysis patients receive medications such as angiotensin converting enzyme (ACE) inhibitors and beta-adrenergic blockers, which may induce or exacerbate symptoms of cough. A dry cough is commonly associated with ACE inhibitors, which are allosteric competitors for the binding sites of angiotensin-converting enzyme in the lungs. Beta-adrenergic antagonists lead to airflow limitation and an exacerbation of asthma in patients with underlying airway hyperresponsiveness (1).

Also, it has been reported that, in PD patients, the presence of dialysate in the peritoneal cavity may interfere with gastric emptying and intestinal motility (2). This observation may explain an increased occurrence of gastroesophageal reflux disease (GERD) in PD patients. In multiple studies, GERD has been shown to be one of the most common causes of cough in all age groups (3).

Despite these risk factors, no systematic assessment has been made of the prevalence of cough in PD patients. Holley and Piraino (4) reported a small group of PD patients with cough and suggested that a directed study examining chronic cough in continuous ambulatory peritoneal dialysis (CAPD) patients, including the potential contributions of pulmonary and esophageal disease, is needed.

In an initial review of 17 charts of PD patients from our center who were referred for respiratory assessment of chronic cough between 1991 and 1998, we found a variety of contributing factors. Often, more than one potential cause of cough per patient was seen. Potential causes included gastroesophageal reflux (4 cases), congestive heart failure (4 cases), asthma (3 cases), pleural effusions (2 cases), chronic obstructive pulmonary disease (4 cases), post-nasal drip (5 cases), and allergic rhinitis (2 cases) (5).

This study aimed to identify the frequency of persistent cough (at least 4 weeks in duration) and associated factors in a large clinical population of PD and HD patients.

Patients and methods

The study was conducted in the Asthma Centre, CAPD unit, and a hemodialysis unit of the Toronto Western and the Toronto General Hospital.

A survey questionnaire [administered by the same individual (FM)] was undertaken among current renal patients. The survey of PD patients was performed at the weekly peritoneal dialysis clinic at the Toronto Western Hospital over the summer of 1998. It involved patients under the care of two nephrologists (DO and JB). The survey of HD patients was conducted at a Toronto General Hospital hemodialysis unit. Patients were asked 27 questions about cough, onset, frequency, duration, smoking history, and symptoms of GERD, based on published surveys of cough and respiratory symptoms (6–8). Furthermore, where possible, clinical charts of these patients were reviewed for medication use since the initiation of dialysis.

Data were analyzed with the Statistical Analysis System statistical package for the PC (SAS Institute, Inc., Cary, NC, U.S.A.). A persistent cough was defined as a cough lasting at least 4 weeks. A total of 92 PD patients and 91 HD patients who answered the question “Have you had the onset of a dry persistent cough since starting dialysis, yes/no” were examined for associations between cough and other responses from the questionnaire and from the renal chart review. Chi-square tests were used to analyze categorical data.


Among PD and HD patients, mean age was similar (59.1 years and 60.1 years, respectively), but significantly more PD patients were female (p = 0.001). Fewer PD patients had ever smoked. Despite the lesser smoking history, a significantly greater number of PD patients reported a persistent cough, 21.7% versus 6.6% (p = 0.003). Reports of sputum production, chest tightness, wheezing, and nocturnal dyspnea were not significantly different between the PD and HD groups. However, more PD patients reported a history of allergy (p = 0.004). Reported prevalence of heartburn was similar between the two groups, but a greater percentage of PD patients used reflux medications (H2 antagonists, proton pump inhibitors, and pro-mobility agents), although the difference in the use of these medications was not statistically significant (70% versus 60%, p = 0.16). Use of ACE inhibitors and beta-blocking medications were similar between the two groups (Table I).

Comparisons of PD patients with and without persistent cough (Table II) showed a significant association between cough and heartburn in these patients (p = 0.008). One third of patients with a persistent cough felt that the cough was affected by dialysis. A significant association was seen between cough and wheezing (p = 0.017), as were statistically nonsignificant trends to associations between cough and sputum production (p = 0.06), nocturnal dyspnea (p = 0.06), and chest tightness (p = 0.07). In contrast, the six HD patients with cough showed no significant associations with the latter variables or others (Data not shown) when compared with HD patients without cough.

For reasons of comparison in the present study, the HD group was included using the same methodology as was used for the PD patient survey. The 22% rate of persistent dry cough in the surveyed PD patients is significantly greater than the 7% rate of cough in the surveyed HD patients. Significantly more PD patients with persistent cough reported heartburn and wheeze.

The frequency of reported cough in 52% of PD patients in this study, and a persistent cough frequency of 22%, cannot be directly compared to cough in the general population. Although a four-week duration of cough was selected to exclude transient cough owing to upper respiratory viral infections (3), prevalence rates of persistent dry cough using this definition have not been reported for the general population. A British study (6) found that the reported rate of “cough every day or on half the days of the year” for people aged 35 and greater was 14.45%. The frequency of a usual morning cough for the same group was 15.7%. In another survey of individuals with a mean age of 44.9 years, which was repeated eight years later, the rate of presence of cough on most days for three months of the year, for two years or more, was 17.9% and 16.5% (7).

Consideration should be given to the fact that self-reporting of cough has been shown to be inaccurate, in relation to the use of portable cough recording devices (8). Our results for both PD and HD patients may underestimate the true prevalence of cough, as may other questionnaire studies.

Gastroesophageal reflux disease (GERD) is common in the general population and is a commonly implicated cause of a chronic persistent dry cough (9). Irwin et al found that GERD caused cough in 40% of patients with chronic cough, and was present in 34% – 89% of asthmatics, although 43% of their patients with GERD-related cough denied heartburn or a sour taste in the mouth (10). An epidemiologic study performed in the 1970s suggested that 10% of the U.S. population have daily heartburn and that more than one third have intermittent symptoms (10). The frequency of heartburn among our surveyed PD patients was about 37%, not dissimilar to reports of intermittent heartburn in the general population. However, the frequency of patients on reflux medication was substantially higher at 70%.

Because intra-abdominal pressures are increased by peritoneal dialysis, PD patients may be expected to tend to experience reflux. Upper gastrointestinal symptoms are often reported by patients undergoing peritoneal dialysis, especially CAPD therapy (11). Reflux may also relate to the frequent subclinical complication of esophageal dysmotility in chronic renal failure (12,13). Gastroesophageal dysmotility, combined with increases in intra-abdominal pressure during PD, is likely to result in GERD and symptoms of reflux such as heartburn and chronic cough.

Other factors may also have contributed to chronic cough in PD patients in this study, as illustrated by our initial review of referred dialysis patients with cough. Pulmonary edema may have caused cough in some patients (14–16). Although CAPD produces symptomatic improvement in patients with refractory congestive heart failure (17), chronic congestive heart failure may be exacerbated by dialysis owing to inadequate fluid removal (18). Similar to findings in the literature, roughly 90% of the patients on dialysis in our study had hypertension documented in their charts, and ACE inhibitors and beta-blockers were commonly prescribed. These factors may have caused or contributed to cough in some patients. The increased self-reported history of allergy in our PD group of patients was not objectively assessed by allergy skin testing or serum immunoglobulin E (IgE) antibody levels. A detailed history concerning the presence of allergy was not available in these patients, and it is possible that they may have incorrectly interpreted the presence of a dry persistent cough as an allergy manifestation. More PD patients than HD patients had never smoked, and smoking cannot therefore be considered as the explanation for cough in this group.


Our findings of an increase in reported chronic cough and an association between chronic cough and GERD in PD patients suggests that additional management measures for GERD should be considered for these patients. Intraperitoneal volumes should perhaps be reduced for patients with chronic cough during periods of frequent coughing (19). Further studies may be helpful to assess such interventions. Other factors contributing to cough—such as post-nasal drip, asthma, chronic obstructive pulmonary disease, pulmonary edema, and use of ACE inhibitors or beta-blocking medications—should also be considered in individual patients.


This study was funded in part by the Ontario Thoracic Society. We thank the nurses of the Home Peritoneal Dialysis Unit for their assistance in the survey. The statistical analysis was greatly facilitated by the efforts of Justina Greene from the Gage Occupational and Environmental Health Unit.


Corresponding author:
Susan M. Tarlo, mb, bs, Toronto Western Hospital, EC4-009, 399 Bathurst Street, Toronto, Ontario M5T 2S8 Canada.From: The University Health Network, Toronto, Ontario, Canada.