Gastro-esophageal reflux disease and exacerbations in chronic obstructive pulmonary disease
Abstract
Background and objective
We tested the hypothesis that gastro-esophageal reflux disease is a risk factor for exacerbations in individuals with chronic obstructive pulmonary disease (COPD).
Methods
Among 9622 participants in the Copenhagen City Heart Study, we identified 1259 individuals with COPD and information on gastro-esophageal reflux disease and the regular use of acid inhibitory treatment. These individuals were followed for 5 years with regard to medically treated COPD exacerbations, which we defined as a short course treatment with oral corticosteroids alone or in combination with antibiotics. We applied a multivariable Cox regression analysis with adjustment for well-established risk factors associated with COPD exacerbations or gastro-esophageal reflux disease, including COPD severity, and symptoms.
Results
Individuals with COPD and gastro-esophageal reflux disease had more chronic bronchitis (31 vs 21%, P = 0.004), more breathlessness (39 vs 22%, P < 0.001), and more of them had a history of respiratory infections (6.8 vs 1.4%, P < 0.001) than individuals with COPD but without gastro-esophageal reflux disease. Among individuals with COPD and gastro-esophageal reflux disease, those who did not use acid inhibitory treatment regularly had an increased risk of COPD exacerbations during follow-up, hazards ratio (HR): HR = 2.7 (1.3–5.4, P = 0.006). Individuals with gastro-esophageal reflux disease, using acid inhibitory treatment regularly did not have an increased risk of exacerbations, HR = 1.2 (0.6–2.7, P = 0.63).
Conclusions
Gastro-esophageal reflux disease was associated with an increased risk of medically treated exacerbations of COPD, but only in those individuals who did not use acid inhibitory treatment regularly.
Abbreviations
-
- CCHS
-
- Copenhagen City Heart Study
-
- COPD
-
- Chronic Obstructive Pulmonary Disease
-
- ECLIPSE
-
- Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints
-
- FEV1
-
- forced expiratory volume in 1 s
-
- FVC
-
- forced vital capacity
-
- GERD
-
- gastro-esophageal reflux disease
-
- GOLD
-
- Global Initiative of Obstructive Lung Disease
-
- HR
-
- hazards ratio
-
- NSAID
-
- non-steroidal anti-inflammatory drug
Introduction
Chronic obstructive pulmonary disease (COPD) is one of the most prevalent global health problems.1, 2 COPD is characterized by the presence of airflow limitation, breathlessness, coughing, sputum and exacerbations.3 Exacerbations are frequent in some individuals with COPD4 and constitute key elements in assessment, prognosis and choice of treatment.5-8
During the last decade, several studies have suggested an association between airway pathology and gastro-esophageal reflux disease (GERD),9 and there is a growing body of evidence indicating an increased prevalence of GERD in COPD.10-13 Nevertheless, to our knowledge, only few studies have prospectively investigated the association between GERD in COPD and exacerbations, and few studies have addressed the role of acid inhibitory treatment in this association.4, 14, 15
We therefore tested the hypothesis that GERD is a risk factor for exacerbations in individuals with COPD.
Methods
Study populations
In this study, we had access to data from 9622 participants who had answered a questionnaire and had performed pulmonary function tests as part of the 1991–1994 examination of the Copenhagen City Heart Study (CCHS).16 This study was approved by an institutional review board and the Danish ethics committee (No. KF-100.2039/91) and was conducted according to the Declaration of Helsinki. Written informed consent was obtained from all participants including participants' consent to collect further information on health issues. An additional permission to merge data from the Copenhagen City Heart Study with data on medication use has been given by the Danish Health and Medicines Authority in 2005 (Journal number 5121-59). Among these participants, for this study, we identified a subgroup of 1279 (13.3%) individuals with COPD defined by a forced expiratory volume in 1 s (FEV1) divided by forced vital capacity (FVC), FEV1/FVC < 0.7, age above 40 years, and no self-reported asthma. Among these individuals with COPD, 1259 (98.4%) had answered questions on presence or absence of night-time and daytime GERD. The exact questions used to define these conditions were: ‘Do you experience heartburn during night-time?’ and ‘Do you experience heartburn during daytime?’ Reporting night-time and daytime GERD is likely to represent the same disease, and it has been shown that particularly reporting night-time heartburn in addition to more unspecific clinical symptoms, such as pain or discomfort in the chest, significantly increases the validity of identifying esophageal disease.17 We therefore based our main clinical definition of GERD on reporting coexisting night-time and daytime GERD.
Using complete record linkage to the national Danish Registry of Medicinal Products Statistics,18 we identified all prescriptions of oral corticosteroids and antibiotics. Clusters of oral corticosteroids, with or without antibiotics, dispensed less than 4 weeks apart, were used to define medically treated exacerbations of COPD.7, 8, 19 The national Danish Registry of Medicinal Products Statistics was established in 1995 and the information is thus available from 1 January 1995. Therefore, the first individual examined in the 1991–1994 examination of the CCHS on 10 October 1991 was followed-up for 5 years, starting from 1 January 1995. This time difference of 1179 days between 10 October 1991 and the index date 1 January 1995 was added to the starting point in time for the follow-up of all individuals. This way, the last individual examined in the CCHS on 29 June 1994 was followed-up from 20 September 1997, until 20 September 2002, thus ensuring that all individuals had an equal follow-up of 5 years and an equal shift in time frame of 1179 days from their examination date.
Statistical analysis
Demographics
We used the statistical software package R (version 3.0.1) in all analyses.20, 21 Characteristics of individuals with COPD reporting coexisting night-time and daytime GERD symptoms, either night-time or daytime GERD but not coexisting, night-time GERD, or daytime GERD at the examination were compared using chi-square tests or ANOVA for categorical or continuous variables as appropriate.22
Gastro-esophageal reflux and medically treated exacerbations
Univariable and multivariable Cox regression analyses22 were applied to explore the association of coexisting night-time and daytime GERD at the examination, regular use of acid inhibitory treatment and time to the first medically treated exacerbation of COPD for each individual during the 5-year follow-up. Among the 1259 individuals with data on coexisting night-time and daytime GERD, 148 (11.8%) died before the index date of data access in the national Danish Registry of Medicinal Products Statistics and were excluded from these analyses. During follow-up, among the remaining 1111 individuals, censoring was death (n = 310) or end of follow-up (5 years).
As our main analysis, we analyzed GERD defined as reporting coexisting night-time and daytime GERD, with or without a regular use of acid inhibitory treatment. The regular use of acid inhibitory treatment was defined by reporting a daily or almost daily use of acid inhibitory treatment. Furthermore, since we only had pre-bronchodilator values, we did a subgroup analysis, where we included only individuals with Global Initiative of Obstructive Lung Disease3 (GOLD) 2–4 grade airflow limitation. In addition, we did two sensitivity analyses with night-time GERD, or daytime GERD, as independent variables of interest.
In multivariable analyses, we included confounders associated with GERD and/or exacerbations of COPD according to previous literature: age, gender, GOLD 1–4 grade, body mass index, smoking, symptoms (breathlessness, chronic bronchitis, wheezing, dysphagia), history of respiratory infections and regular use of respiratory medications.3, 23-25 Table S1 in the online supporting information shows a detailed description of all variables. To adjust for confounding by a possible healthy user effect,26 we also adjusted our analysis for regular use of vitamin pills defined by reporting a daily or almost daily use of vitamin pills. Furthermore, because arthritis medication, which include non-steroidal anti-inflammatory drugs (NSAID), is associated with gastro-esophageal reflux and gastric ulcers,27 we also adjusted our analysis for reported regular use of arthritis medications; see Table S1 in the online supporting information for details.
Attributable risk
From the prevalence of individuals with coexisting night-time and daytime GERD with exacerbations, we used the relative risk between those with regular acid inhibitory treatment and those without to calculate the percentage of exacerbations that may be attributed to the lack of acid inhibitory treatment among individuals with coexisting night-time and daytime GERD.28, 29
Results
Demographics
Among the 1259 individuals with COPD and information on night-time and daytime GERD, 74 (5.9%) reported coexisting night-time and daytime GERD at the examination, 112 (8.9%) reported night-time GERD and 124 (9.8%) individuals reported daytime GERD. Table 1 shows the demographics of the study population according to the presence of coexisting night-time and daytime GERD, either night-time or daytime GERD but not coexisting, and no GERD, at the examination in 1991–1994. Table S2 in the online supporting information shows the characteristics of individuals reporting night-time GERD or daytime GERD. Individuals with COPD and GERD reported more respiratory symptoms and dysphagia. As expected, individuals with GERD had more often a regular use of acid inhibitory treatment than those without GERD; around half of those reporting coexisting night-time and daytime GERD were on regular treatment (Table 1), and 20% of those reporting either night-time or daytime GERD but not coexisting were on regular treatment. From the examination group, 148 (11.8%) individuals died before the index date (1 January 1995). Among the remaining 1111 individuals, 66 (5.9%) reported coexisting night-time and daytime GERD, 100 (9.0%) reported night-time GERD and 111 (10.0%) reported daytime GERD.
| Variables | No GERD | Either night-time or daytime GERD, but not coexisting | Coexisting night-time and daytime GERD | P-value |
|---|---|---|---|---|
| n = 1097 (87.1%) | n = 88 (7.0%) | n = 74 (5.9%) | ||
| Age (mean (SD)) | 66.9(9.7) | 67.3(9.5) | 67.8(10.5) | 0.71 |
| BMI ≥ 25(%(No.)) | 42.8(469) | 46.6(41) | 52.7(39) | 0.21 |
| GOLD 1(%(No.)) | 32.1(352) | 37.5(33) | 36.5(27) | 0.33* |
| GOLD 2(%(No.)) | 54.0(592) | 45.5(40) | 55.4(41) | |
| GOLD 3+4(%(No.)) | 13.9(153) | 17.0(15) | 8.1(6) | |
| Men(%(No.)) | 52.9(580) | 47.7(42) | 64.9(48) | 0.08 |
| Chronic bronchitis(%(No.)) | 21.2(231) | 34.1(30) | 31.1(23) | 0.004 |
| Breathlessnessa(%(No.)) | 22.1(242) | 33.0(29) | 39.2(29) | <0.001 |
| History of respiratory infectionsb(%(No.)) | 1.4(15) | 8.0(7) | 6.8(5) | <0.001h |
| Wheezing(%(No.)) | 36.8(403) | 44.8(39) | 44.6(33) | 0.15 |
| Current smoking(%(No.)) | 69.2(759) | 68.2(60) | 55.4(41) | 0.05 |
| Former smoking(%(No.)) | 21.2(233) | 21.6(19) | 39.2(29) | 0.002 |
| Dysphagiac(%(No.)) | 2.3(25) | 12.5(11) | 16.2(12) | <0.001 |
| Regular use of acid inhibitory treatmentd(%(No.)) | 3.7(40) | 20.9(18) | 56.2(41) | <0.001 |
| Regular use of pulmonary medicatione(%(No.)) | 5.0(54) | 4.7(4) | 4.2(3) | 1h |
| Regular use of vitamin pillsf(%(No.)) | 59.3(644) | 62.1(54) | 61.1(44) | 0.85 |
| Regular use of arthritis medicationsg(%(No.)) | 6.3(69) | 11.6(10) | 13.7(10) | 0.02 |
- *This P-value compares all GOLD groups.
- a A score on the modified Medical Research Council scale(mMRC) ≥ 2.
- b Reports at least six respiratory infections during the last 10 years before the examination that required the individual to stay home from work and/or consult a doctor.
- c Reports difficulty swallowing liquids or solid food.
- d Reports daily or almost daily use of acid inhibitory treatment.
- e Reports daily or almost daily use of pulmonary medication.
- f Reports daily or almost daily use of arthritis medications.
- g Reports daily or almost daily use of vitamin pills.
- h Fisher's exact test.
- This table shows the examination demographics for 1259 individuals with COPD in the 1991–1994 examination of the Copenhagen City Heart Study, according to status of GERD at the examination.
- Please note that some of the variables do not add up to a total of 1259 due to a few missing data.
- BMI, body mass index; COPD, chronic obstructive pulmonary disease; GERD, gastro-esophageal reflux disease; GOLD, Global Initiative of Obstructive Lung Disease; SD, standard deviation.
GERD and medically treated exacerbations
During the 5-year follow-up, we analyzed 185 medically treated exacerbations included as outcome variable in Cox regression analyses.
In univariable analysis, coexisting night-time and daytime GERD was associated with medically treated exacerbations in COPD among individuals not using acid inhibitory treatment regularly, hazards ratio (HR) = 2.1 (1.1–4.1, P = 0.03), whereas no association was observed among regular users of acid inhibitory treatment, HR = 1.2 (0.6–2.5, P = 0.65).
As shown in Table 2, in multivariable analysis, coexisting night-time and daytime GERD was associated with exacerbations in COPD among the individuals not using acid inhibitory treatment regularly, HR = 2.7 (1.3–5.4, P = 0.006), but not among regular users of acid inhibitory treatment, HR = 1.2 (0.6–2.7, P = 0.63) (P = 0.03 in test for interaction between coexisting GERD and regular use of acid inhibitory treatment).
| Variables | Risk of exacerbations | Subgroup analysis: GOLD 2–4 | ||
|---|---|---|---|---|
| HR (95% CI) | P-value | HR (95% CI) | P-value | |
| Coexisting night-time and daytime gastro-esophageal reflux disease and no regular use of acid inhibitory treatmenta | 2.7(1.3–5.4) | 0.006 | 2.6(1.2–5.8) | 0.02 |
| Coexisting night-time and daytime gastro-esophageal reflux disease and regular use of acid inhibitory treatmenta | 1.2(0.6–2.7) | 0.63 | 0.8(0.3–2.3) | 0.68 |
| Either night-time or daytime gastro-esophageal reflux disease but not coexisting, and no regular use of acid inhibitory treatmenta | 1.7(1.0–3.0) | 0.05 | 1.2(0.6–2.5) | 0.55 |
| Either night-time or daytime gastro-esophageal reflux disease but not coexisting, and regular use of acid inhibitory treatmenta | 0.3(0.05–2.4) | 0.28 | 0.4(0.05–2.6) | 0.32 |
| No gastro-esophageal reflux disease but regular use of acid inhibitory treatmenta | 1.8(0.9–3.5) | 0.08 | 1.4(0.6–3.2) | 0.46 |
| Men | 0.6(0.4–0.8) | <0.001 | 0.7(0.5–0.9) | 0.02 |
| Age (per 10 year increase) | 1.2(1.05–1.5) | 0.01 | 1.2(1.0–1.5) | 0.08 |
| BMI ≥ 25b | 0.8(0.6–1.1) | 0.13 | 0.8(0.5–1.1) | 0.13 |
| Dysphagia | 0.8(0.4–1.8) | 0.65 | 1.0(0.4–2.4) | 0.98 |
| GOLD 2c | 1.7(1.2–2.5) | 0.004 | 1.0(Reference) | NR |
| GOLD 3 + 4 | 3.3(2.1–5.3) | <0.001 | 1.8(1.2–2.7) | 0.003 |
| Wheezing | 1.0(0.7–1.4) | 0.83 | 1.1(0.7–1.5) | 0.78 |
| Breathlessnessd | 1.0(0.7–1.5) | 0.85 | 1.2(0.8–1.8) | 0.35 |
| History of respiratory infectionse | 1.2(0.6–2.2) | 0.64 | 1.3(0.6–2.6) | 0.48 |
| Chronic bronchitis | 1.3(0.9–1.9) | 0.13 | 1.2(0.8–1.8) | 0.29 |
| Current smoking | 1.7(1.0–3.1) | 0.07 | 2.9(1.2–6.7) | 0.02 |
| Former smoking | 1.1(0.6–2.0) | 0.87 | 1.6(0.7–4.1) | 0.29 |
| Regular use of pulmonary medicationf | 1.7(1.0–3.1) | 0.07 | 1.5(0.8–2.8) | 0.22 |
| Regular use of vitamin pillsg | 1.2(0.8–1.6) | 0.34 | 1.2(0.8–1.7) | 0.31 |
| Regular use of arthritis medicationsh | 1.7(1.1–2.9) | 0.03 | 1.1(0.6–2.2) | 0.76 |
- a Reference group: no gastro-esophageal reflux disease and no regular use of acid inhibitory treatment.
- b Reference group: BMI < 25.
- c Reference group: GOLD 1 grade COPD.
- d Breathlessness = a score on the Modified Medical Research Council scale, mMRC ≥ 2.
- e Reports at least six respiratory infections within the last 10 years before the examination requiring consulting a doctor and/or staying home from work.
- f Reports daily or almost daily use of pulmonary medication.
- g Reports daily or almost daily use of vitamin pills.
- h Reports daily or almost daily use of medicine against arthritis.
- Multivariable Cox regression analysis: shows risk of having medically treated exacerbations among 1084 individuals with COPD in the 1991–1994 examination of the Copenhagen City Heart Study who had complete data on all variables included. A total of 185 exacerbations during follow-up were recorded.
- BMI, body mass index; CI, confidence interval; COPD, chronic obstructive pulmonary disease; GOLD, Global Initiative of Obstructive Lung Disease; HR, hazards ratio; NR, not relevant.
The results of the subgroup analysis, including only individuals with GOLD 2–4 grade airflow limitation, were similar to the main results, as shown in Table 2. Sensitivity analyses, among individuals with nighttime GERD or daytime GERD as variables of main interest also resulted in similar results, confirming GERD as a strong risk factor of exacerbation among those individuals not using acid inhibitory treatment regularly, but not among regular users of acid inhibitory treatment (P = 0.03 in test for interaction between night-time GERD and regular use of acid inhibitory treatment, and P = 0.04 in test for interaction between daytime GERD and regular use of acid inhibitory treatment), as shown in Tables S3 and S4 in the online supporting information.
Confounding variables
The relationships between confounding variables and risk of exacerbations are shown in Table 2 and Tables S3 and S4 in the online supporting information. As expected, older age and poor lung function (here included as GOLD 1–4 grade airflow limitation) were associated with an increased risk of exacerbations. In addition, males had a lower risk than females, whereas the regular use of respiratory medications, smoking and the presence of bronchitis were less strongly associated with exacerbations. Finally, we also observed an association between the regular use of arthritis medications and the increased risk of COPD exacerbations.
Attributable risk
We estimated that among individuals with COPD exacerbations and coexisting night-time and daytime GERD, approximately 31% of their exacerbations could be attributed to lack of regular acid inhibitory treatment.
Discussion
We found, that GERD is associated with COPD exacerbations, but only among individuals not using acid inhibitory treatment regularly. Our findings are based on the first prospective analyses to our knowledge of GERD, the regular use of acid inhibitory treatment and the medically treated exacerbations in individuals with COPD from the general population.
Few earlier studies have tried to access GERD as a risk factor of exacerbations in individuals with COPD.4, 14, 15 Most importantly, the study by Hurst et al. on patients from the Evaluation of COPD Longitudinally to Identify Predictive Surrogate Endpoints (ECLIPSE) cohort showed an approximately 1.7 times higher risk of medically treated exacerbations in patients reporting GERD.4 We believe that our study supports and expands these findings by showing that this increased risk applies only among those individuals not using acid inhibitory treatment regularly. Exacerbations in ECLIPSE were defined as events treated with oral corticosteroids and/or antibiotics.4 We required the use of oral corticosteroids in the definition of exacerbations in our study in line with other previous studies including both randomized controlled trials and population-based studies.8, 30 However, because the use of antibiotics alone could also represent an exacerbation, some exacerbations in our study might have been missed. Nevertheless, in our real-life setting, prescription data could not be used defining COPD exacerbations, because it would be impossible to differentiate between the use of antibiotics for various infections and the use for exacerbations. In addition, it is also important to recognize that although the individuals using acid inhibitory treatment reported being highly adherent to these medications, our exacerbation definition was based on prescription data. This implies that some individuals, although physically retrieving their systemic corticosteroids at the pharmacy, might not have used them. Possible explanations for not using their medication could include a spontaneous reduction in symptoms or perhaps stockpiling of medication.
Mechanistically, an explanation for the association between GERD and exacerbations of COPD could be that the aspiration of gastric acid causes airway inflammation.9 Proton pump inhibitors virtually abolish acid secretion in normal clinical doses.31 We find a null-association between GERD and COPD in users of acid inhibitory treatment, which altogether suggest that the acidity of the reflux content could be the key to the link between reflux and COPD exacerbations. We estimated that among individuals with COPD exacerbations and coexisting night-time and daytime GERD, around one third of their exacerbations could be attributed to lack of regular acid inhibitory treatment. This needs, however, to be investigated in a randomized controlled trial.12 Results from studies using proton-pump inhibitors aiming at lowering the risk of asthma exacerbations,32, 33 or as treatment of poorly controlled asthma, have however been disappointing.34, 35 Furthermore, treatment with regular acid inhibitory treatment in COPD is symptomatic treatment and not without adverse effects. In fact, previous studies showed that the use of proton-pump inhibitors can be associated with an increased risk of pneumonia.36, 37 In addition, the effect of GERD on exacerbations could very well be influenced by other factors such as non-acidic reflux and pepsin.38
Another, in our opinion less likely, explanation linking GERD with exacerbations is that GERD could cause symptoms such as cough that could be perceived as an exacerbation by both patients and doctors, thereby resulting in the treatment with systemic corticosteroids. Treatment for GERD-related cough, using acid inhibitory treatment regularly, may reduce this symptom sufficiently to prevent treatment with systemic corticosteroids. Thus, our observation of an association between GERD and exacerbations could be a reflection of the association between GERD and respiratory symptoms as such. In fact, as in previous studies,10, 24 we observed that individuals with COPD and GERD report significantly more breathlessness, wheezing and chronic bronchitis and have more often a history of respiratory infections than individuals without GERD.
Our data suggest that the risk of exacerbation is also higher among those individuals who regularly use arthritis medications. It is recognized that gastro-esophageal reflux and damage to the gastro-intestinal mucosa are side-effects to a regular use of non-steroidal anti-inflammatory treatment, and furthermore, it is also acknowledged that regular acid inhibitory treatment protects against these side-effects.39 Therefore, we believe that the association between the regular use of arthritis medications and the risk of COPD exacerbation strengthens our main findings.
Further strengths of our study includes linkage to an all-inclusive register containing information on all dispenses of systemic corticosteroids and antibiotics during follow-up. An important limitation to our design, however, was the fact that the registry only holds information from 1 January 1995 and forward. Therefore, we were not able to follow all individuals with regard to medically treated exacerbations, as 148 individuals died before this date. We compensated for this by introducing a window of 1179 days for all participants from the date of examination until the date that was used as the start of the follow-up. Although this procedure ensures each individual an equal follow-up time, and an equal shift in timeframe from their examination date, we cannot rule out that this might have introduced some bias because for example disease severity and presence of comorbidities could change, which we could not take into account in the present study. On the other hand, this artificial window between the examination and the start of follow-up is more likely to weaken rather than strengthen any possible association between baseline variables and exacerbations, and the fact that we found quite robust associations suggests that our findings are not caused by chance.
Assessment of symptoms using questionnaires has been recommended as primary outcome in clinical GERD trials.40 However, a further limitation to our study is our definition of GERD only by the presence of heartburn. Although heartburn is a highly specific question in the diagnosis of GERD, the sensitivity is rather low.41 The use of a validated questionnaire, endoscopy or esophageal pH measurements would provide stronger evidence on the observed association. Nevertheless, the sensitivity of endoscopy is also low, and although esophageal pH testing would be the gold standard, its utility in clinical practice is limited.41 Furthermore, it has been shown that adding the presence of night-time heartburn significantly increases the validity of symptoms of GERD.17
Furthermore, although our study is prospective with respect to COPD exacerbations, it is observational with respect to GERD symptoms and the regular use of acid inhibitory treatment that were recorded at the examination and, as another limitation to our study, it is important to notice that some individuals might not report GERD due to a regular use of acid inhibitory treatment. Therefore, although the observed association was strong, it is difficult to draw conclusions about causality from our design.
Another limitation to our design is a possible healthy user bias; that is, that a person using any kind of medication regularly might have some characteristics that enhances several beneficial factors that could reduce the risk of exacerbations.26 In this study, we adjusted all analyses for a regular use of vitamin pills as a surrogate marker for a healthy user bias because vitamin pills are not a priori expected to affect the risk of exacerbations. This adjustment may not be a sufficient surrogate marker for a possible healthy user bias, and residual confounding is still possible. Nevertheless, this adjustment did not weaken the association among GERD, the regular use of acid inhibitory treatment and the exacerbations, suggesting that a possible healthy user effect is not a major confounder to our findings. Furthermore, we tested if there was any association between a regular use of acid inhibitory treatment and a regular use of pulmonary medication, a group of medication expected to affect the risk of exacerbations, and this was not the case.
We believe that a major strength to our study is our ability to adjust for several variables known to be associated with GERD or exacerbations in COPD and that these adjustments only modestly affected our estimates.
In conclusion, our study shows that GERD is associated with medically treated exacerbations in COPD, but only in those individuals not using acid inhibitory treatment regularly. Our findings indicate that a randomized trial is merited to explore the effects of the different types of acid inhibitory treatment on exacerbations in COPD.
Acknowledgements
This study was supported by the Capital Region of Copenhagen, Danish Heart Foundation, Danish Lung Foundation, Velux Foundation and Herlev Hospital. The funding sources had no role in the design and conduct of the study; in the collection, management, analysis and interpretation of the data; nor in the preparation, review or approval of the manuscript.
