The Association Between Proton Pump Inhibitors and Incident Asthma in Patients with Coronary Artery Disease: A Population-Based Cohort Study

Observational studies have found a signicant association between acid-suppressive drug use and incident asthma. However, the association between proton pump inhibitors (PPIs) and incident asthma in patients with coronary artery disease (CAD) is unclear. Thus, this study assessed the association between PPI use and incident asthma in patients with CAD. We conducted a retrospective cohort study using the National Health Insurance Research Database in Taiwan from 2004 to 2013. Each patient who took PPIs was assigned to the PPI group, whereas 1:1 sex-, age-, and drug index date-matched randomly selected patients without PPI prescription were assigned to the non-PPI group. We analyzed the risk of incident asthma using Cox proportional hazard regression models, including sex, age, urbanization, low income, and comorbidities. Sensitivity and subgroup analyses were also conducted. A total of 8499 cases were identied as PPI ‐ treated patients, and 8,499 subjects were included in the control group of PPI non ‐ users. After adjusting for sex, age, urbanization, low income, and comorbidities, PPI user was associated with a 1.18-fold (HR: 1.18; 95% CI: 1.05–1.34) increase for incident asthma in patients with CAD. We concluded that PPI use increased the risk of incident asthma in patients with CAD. The risk of incident asthma should be considered when weighing the benets and risks of PPI and aspirin treatment in patients with CAD in clinical practice. reux measurable Library systematic the of several antireux on outcomes in children and adults review, 2009, studied whether or not treatment of GERD with PPIs in children . Both studies were limited by small-scale RCTs using PPIs with conicting results and provided no denitive recommendations regarding the use of PPIs for The meta-analysis showed that prenatal exposure to drugs, PPIs,


Introduction
Asthma is a common disease among children. 1 In 2016, WHO estimated 417,918 deaths due to asthma at the global level and 24.8 million disability-adjusted-life-years attributable to asthma 2 . Numerous comorbidities can be associated with asthma and in uence its clinical expression, but their speci c in uence remains to be characterized. Frequently contributing comorbid conditions reported in asthmatic patients include rhinitis, sinusitis, gastroesophageal re ux disease (GERD), obstructive sleep apnea, hormonal disorders, and psychopathologies, although other conditions, sometimes without an evident link with asthma, are also highly prevalent in asthmatic patients 3 .
Several studies have investigated the e cacy of different PPIs on asthma outcomes through randomized controlled trials (RCTs) [4][5][6][7][8] . Some studies have indicated that symptoms, lung function, or both can be improved with the treatment of acid re ux [7][8][9] ; others have not demonstrated measurable improvement with acid suppression 4,10,11 . A Cochrane Library systematic review published in 2003 examined the effects of several antire ux treatments on asthma outcomes in children and adults 12 . The most recent systematic review, published in 2009, studied whether or not treatment of GERD with PPIs improves asthma symptoms in children 13 . Both studies were limited by small-scale RCTs using PPIs with con icting results and provided no de nitive recommendations regarding the use of PPIs for patients with asthma. The meta-analysis showed that prenatal exposure to acid-suppressive drugs, such as histamine H2-receptor antagonists and PPIs, is associated with an increased risk of childhood asthma.
The evidence suggests that prenatal, maternal, acid-suppressive drug use is associated with an increased risk of childhood asthma 14 .
Coronary artery disease (CAD) is one of the major cardiovascular diseases affecting the global human population 15 . Daily low-dose aspirin is recommended for the prevention of cardiovascular events in patients with CAD, and PPI is recommended to prevent or treat aspirin-associated gastrointestinal injury [16][17][18] . However, the association between PPIs use and incident asthma in patients with CAD remains unclear till now. Therefore, we examined the risk for incident asthma associated with PPI use in patients with CAD in the National Health Insurance Research Database (NHIRD) of Taiwan.

Results
Demographics. We identi ed 25,082 CAD patients with new diagnosed peptic ulcer or GERD, of whom we excluded 2312 who had asthma before 2004 and 637 dead during the follow-up period. Thus, we assessed data for 8,499 PPI users and 13,504 non-PPI users. After 1:1 age, sex, and drug index date, 8,499 non-PPI users were used for analyzing the relationship between the use of PPI and incident asthma ( Figure 1). Characteristics at baseline are shown in Table 1 The association between PPI use and incident asthma. Table 2 shows 706 asthma events (387 for the PPI cohort and 319 for the non-PPI cohort). The incidence was 1.21-fold higher (95% CI, 1.06-1.36) in the PPI cohort than in the non-PPI cohort (5.38 vs. 4.45 per 10,000 person-months), with an adjusted HR of 1.18 (95% CI, 1.05-1.34) after controlling for sex, age, urbanization, low income, and comorbidities.
Kaplan-Meier method analysis showed that the cumulative probability of developing asthma from 2004 to 2013 follow-up period showed signi cant differences between the PPI and non-PPI cohorts (log-rank test, p < 0.0001; Figure 2).
Sensitivity and subgroup analysis. Sensitivity analysis for the multiple Cox regression model con rmed the risk of incident asthma in the patients with PPI use when assessed using IPTW and propensity score matching (Table 3). In the subgroup analysis, the risk of asthma was signi cantly greater in the patients with male and female with PPI use than in patients without PPI use (aHR: 1.20, 95% CI: 1.06-1.41 versus aHR: 1.15, 95% CI: 1.01-1.36, respectively) ( Table 4). However, the differences between these comorbidities were not differ among patients with hypertension, diabetes mellitus, hyperlipidemia, stroke, urticaria, atopic dermatitis, allergic rhinitis, pneumonia, cancer, or depression (Table 4).

Discussion
This large population-based cohort study found PPI use increased a 1.18-fold risk of asthma development in patients with CAD. However, subgroup analysis revealed no association between PPI use and the risk of incident asthma compared with non-PPI use in patients with hypertension, diabetes mellitus, hyperlipidemia, stroke, urticaria, atopic dermatitis, allergic rhinitis, pneumonia, cancer, or depression.
A signi cantly higher prevalence of asthma has been previously reported in patients with PPI use as compared with patients without PPI use, particularly in children [19][20][21][22] . Similarly, a recent study in children after high-dimensional propensity score matching has revealed a higher risk of incident asthma with PPI use comparable with non-PPI use (HR, 1.48; 95% CI, 1.41-1.55) 22 . Wang et al. conducted a population-based registry study in Sweden, which included 17,740 children with follow-up between January 2007 and June 2016. 22 They used data from nationwide Swedish registries, with health care and administrative records on both PPIs and registry-recorded hospital diagnoses of asthma. The study reported a 57% increased risk of incident asthma in children exposed to PPI medications. Although the data did not include primary care clinics for the diagnosis of asthma, the study ndings are signi cant and similar to those obtained in our patients with CAD. The biological mechanisms behind the causal relation between PPI use and incident asthma are unclear. PPI medications have been suggested to interfere with the normal digestion of peptides in the stomach, resulting in a T helper 2 dominant response [23][24][25] . This response is thought to be caused by the preservation of epitopes that are normally degraded by exposure to the acidic environment in the stomach. Alternatively, PPI might directly damage the endothelial function and accelerate endothelial senescence of the lung 26 . This endocrine disruption hypothesis could explain the current nding of an association of PPI use with an increased risk of asthma.
Daily low-dose aspirin is recommended for the prevention of cardiovascular events in patients with CAD, and PPI is recommended to prevent or treat aspirin-associated gastrointestinal injury [16][17][18] . Previous studies have reported that low-dose aspirin therapy protects against incident asthma [27][28][29] . However, the current study suggests that PPI treatment increases the risk of asthma development. Therefore, risk of incident asthma should be considered when weighing the bene ts and risks of PPI and aspirin treatment in patients with CAD in clinical practice.
Our study is the largest population-based study to examine the association between PPI use and subsequent asthma development. The national database we used contains a representative cohort of 1,000,000 people covered by the Taiwan NHI program, and the 10-year observation period ensured the power of our statistical analyses. In addition, this is the rst population-based study to suggest that PPI use is associated with an increased risk of asthma development in patients with CAD, although some risk factors for PPI are associated with asthma development.
This study has some limitations. First, all patients in this study were collected from claim datasets of Taiwan NHI that had been submitted from primary care clinics. Risk factors for asthma, such as body mass index, smoking status, family history, treatment adherence, environmental tobacco smoke exposure, and diet, were not available from these secondary data. However, considering that we used population-based data, we assumed that no differences exist between PPI users and non-PPI users. Second, all patients in our study sample had received a diagnosis of CAD and peptic ulcer, and received PPI treatment. However, dosing, treatment adherence, and severity of CAD with acid-related gastrointestinal tract disorders may have differed across patients who used different PPIs. Thus, the association between PPI therapy and risk for asthma may not re ect the effect of prescribed drugs but rather the severity of the patient's diseases and treatment adherence.

Conclusion
In conclusion, this large population-based cohort study showed PPI use increased the relative risk of incident asthma in patients with CAD. However, there are no association between PPI use and the risk of incident asthma compared with non-PPI use in patients with hypertension, diabetes mellitus, hyperlipidemia, stroke, urticaria, atopic dermatitis, allergic rhinitis, pneumonia, cancer, or depression. To select the control group, we used 1:1 matching by sex-, age-, and drug index date in subjects who were not PPI prescription. The drug index date of the PPI was de ned as the date of the rst prescription of PPI in the study period. The endpoint was the development of asthma, which was de ned by the time an asthma (ICD-9 CM code: 493) code rst appeared in the outpatient claim records. Study population. Patients at least one of the following inclusion criteria had to be met: (1) two or more outpatient visits within 6 months, (2) all prescriptions of PPI were continuously administered to the patients for more than 3 months within a 10-year follow-up period, or (3) one or more outpatients with a diagnosis of peptic ulcer disease or GERD. Patients who ful lled any of the following criteria were excluded from the study: (1) had a prior history of asthma before the index date and (2) patients dead within a 10-year follow-up period. Finally, the study group comprised 8,499 participants who were PPI users, and the control group included 8,499 participants who were not PPI users (Figure 1).
Statistical analysis. Data are presented as proportions for the categorical variables. Comparisons between two groups were made using the chi-squared test. The Cox proportional hazard regression model was used to compare the development risk of asthma between PPI and non-PPI users. Adjusted HRs and 95% CIs were calculated. Their values were adjusted for important risk factors, such as sex, age, urbanization, low income, and comorbidities for asthma development. A Kaplan-Meier plot without covariance correction is presented to analyze the risk of incident asthma according to the presence or absence of asthma. In additional, we conducted a sensitivity analyses to test the robustness of our ndings. The propensity score matching (PSMATCH) and inverse probability of treatment weighting (IPTW) was conducted between the two groups by using the SAS software. Finally, we conducted subgroup analyses strati ed by sex, age, urbanization, and comorbidities at baseline for the outcomes of incident asthma respectively. All effects were analyzed using an intention-to-treat approach. Statistical signi cance was considered at P < 0.05. All statistical calculations were performed using statistical analysis software, version 9.3 (SAS Institute, Inc., Cary, NC, USA).    Table 4 Subgroup analysis between PPI and non-PPI groups.  Kaplan-Meier analysis for cumulative incidence of incident asthma between PPI user and PPI non-user group.