Long-Term Proton Pump Inhibitors (PPI) Therapy and Risk for Community-Acquired Pneumonia (CAP): A Systematic Review and Meta-Analysis.

Long-Term Proton Pump Inhibitors (PPI) Therapy and Risk for Community-Acquired Pneumonia (CAP): A Systematic Review and Meta-Analysis.

Toke Nilesh1, Rathod Ajit2, Phalak Pooja*3

1.TOKE NILESH (TN) – Consultant; Department of Gastroenterology & Hepatology, Marengo CIMS Hospital, Ahmedabad, India.

2.RATHOD AJIT (RA) – Specialist; Department of Family Medicine, Al Shahaniya Health Center, Ministry of Public Health, Doha, Qatar.

3.PHALAK POOJA (PP) - Senior Resident; Dept. of OncoPathology, The Gujarat Cancer and Research Institute, Ahmedabad, India.

*Correspondence to: Dr. Phalak Pooja, Senior Resident; Department of OncoPathology, Gujarat Cancer and Research Institute, New Civil Hospital campus, Ahmedabad-380016.


© 2023 Dr. Phalak Pooja. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Received: 18 November 2023

Published: 30 November 2023


Background: Recent studies have suggested that the use of proton-pump inhibitors (PPIs) may increase the risk for community-acquired pneumonia (CAP).

Objective: This review aims to examine the effect of PPI therapy on CAP risk, with a special look at long- term PPI use.

Methods: A systematic literature search was conducted for articles published up to 1st October 2023. The databases used were PubMed, Scopus, ScienceDirect, and Google Scholar. STATA software version 15 was used for statistical analysis using a common-effect inverse-variance model and a p-value of .05 as the significance threshold.

Results: The initial search identified 760 articles. After a study selection process, 15 studies were included in the review and 9 studies in the meta-analysis. This paper included 108, 176 CAP cases and 1,248,785 healthy controls. The adjusted effect sizes for current PPI use across included studies ranged from 1.02 to 3.3, with most of them between 1.02 and 1.5. The meta-analysis performed in this review also found an increased risk for CAP among current PPI users, with an adjusted odds ratio (AOR) of 1.13 (95% CI, 1.09 - 1.18, I2 = 88.8%, p<0.001). These results show that PPI use increases the risk of CAP incidence. Additionally, study findings showed a general trend where the risk of community- acquired pneumonia decreased with increasing duration of PPI use. Short-duration PPI use (<1 year) had higher odds for CAP risk compared to long-term PPI use (>1 year).

Conclusion: Proton-pump inhibitor therapy is associated with an increased risk for CAP. The strength of this association is also found to decrease with increasing duration of PPI use. This means that long- term use of PPI is associated with a relatively lower risk of CAP.

Keywords: Long-term, Proton Pump Inhibitors (PPI), Community Acquired Pneumonia (CAP), Adverse effects.

Long-Term Proton Pump Inhibitors (PPI) Therapy and Risk for Community-Acquired Pneumonia (CAP): A Systematic Review and Meta-Analysis.


Proton pump inhibitors (PPIs) were first used in clinical practice in 1988 and have since been the standard for management of a variety of acid-related gastrointestinal illnesses, including gastro-esophageal reflux disease (GERD), [1, 2, 3] peptic ulcers [4, 5] and non-ulcer dyspepsia. [6, 7] This wide application is due to an accepted safety profile, since they are generally well tolerated by patients, and serious side effects are rarely reported. However, a growing number of publications have linked long-term use of PPIs to adverse effects such as hip fractures, [8] clostridium difficile infection, [9, 10, 11] drug-drug interactions, e.g. clopidogrel [12, 13] and community-acquired respiratory tract infection. [14, 15, 16, 17]

Community-acquired respiratory tract infections are among the most prevalent infectious ailments globally, and they significantly contribute to both mortality and morbidity. A study conducted in the United States in 2008 revealed that, when adjusted for age, the mortality rate attributable to influenza and pneumonia was 20.3 per 100,000 individuals. [18] As of November 2022, the annual incidence of CAP in the United States was 24.8 cases per 10,000 adults with higher rates as age increased. [19] According to Niederman et al. [20] the global mortality for hospitalized CAP averages 12% across regions, but increases in specific populations such as those with bactermia and those from nursing homes. In Regunath and Oba, [19] the mortality rate for those admitted to the intensive care unit due to CAP was as high as 23%. Additionally, Womack and Kropa [21] reported a 30-day mortality for 6% of those hospitalized with CAP.

The pathogens causing CAP are classified as either typical agents such as Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis, Staphylococcus aureus, or as atypical agents such as Legionella, Mycoplasma, Chlamydia pneumoniae, and Chlamydia psittaci. [19] When using sputum culture for CAP etiological diagnosis, the predominant pathogen detected was Streptococcus pneumoniae or pneumococcus, accounting for 9-20% of all CAP cases. [22, 23] Conversely, in cases where serological testing is conducted, Mycoplasma pneumoniae emerges as the most prevalent organism, contributing to 13-37% of the total CAP cases. [22, 23, 24] Additionally, Chlamydia pneumoniae has been reported in about 17% of outpatients diagnosed with CAP. [24] Legionella spp. have also been observed, with rates ranging from 0.7% to 13% across various patient cohorts. [25]

The reported risk factors for the occurrence of CAP are age > 60 years with alcoholism, asthma, heart and lung diseases, immunosuppressive therapy, smoking, low body mass index (BMI), diabetics and history of respiratory infection and pneumonia. [26, 27, 28, 29] Recently, several studies have reported an observed correlation between the use of acid-suppressing medications like PPIs and an increased risk for community-acquired pneumonia (CAP). The exact mechanism by which acid-suppressive medications increase the susceptibility to community-acquired pneumonia (CAP) remains to be fully explained. However, it is postulated that changes in gastric pH, resulting from these medications, cause modifications in the normal microbiota of the gastrointestinal tract and oropharyngeal regions. This potentially leads to diminished pathogen elimination or enhanced pathogen colonization. [29] Specifically, the escalation of gastric pH induced by acid-suppressive agents stimulates the proliferation of microorganisms within the oral and oropharyngeal cavities. [30, 31] Given that gastric acid plays a pivotal role in protecting against infections, the mitigation of its acidic environment presents a plausible mechanism to account for the heightened risk of CAP associated with proton pump inhibitors (PPIs). [32, 33]

This systematic review aims to examine the risk for community-acquired pneumonia caused by PPI use, with a special look at the long-term use of PPIs.



This systematic review is reported following guidelines outlined in the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement. [34]

Information Sources and Study Selection

A systematic literature search was conducted for articles published up to 1st October 2023. The index databases used were PubMed, Scopus, ScienceDirect, and Google Scholar. Topic keywords were used to generate search strings. The identified studies were then subjected to a study selection process.


Table 1: Search strings


Search strings

PubMed/ Scopus/ Google Scholar

(“proton pump inhibitor” OR PPI OR “acid-suppressive” OR “acid suppressant” OR “gastric acid inhibitor” OR “gastric acid suppressor”) AND (“community-acquired pneumonia” OR CAP OR “community pneumonia” OR “outpatient pneumonia”)


(“proton pump inhibitor” OR PPI OR “acid-suppressive” OR “acid suppressant” OR “gastric acid inhibitor”) AND (“community-acquired pneumonia” OR CAP OR “community pneumonia” OR “outpatient pneumonia”)

The search string for ScienceDirect was shorted because the database only accepts search strings with a maximum of 8 Boolean operators. The search in ScienceDirect was limited to the Title, abstract, and keywords, to limit the number of irrelevant studies during searching.


Inclusion and exclusion criteria

For articles to be considered eligible for inclusion, they had to be original research articles and written in the English language. The study population had to be >18 years of age. This review included both observational studies that evaluated PPI use and CAP incidence in human subjects.

Exclusion criteria were non-original research articles like systematic reviews, meta-analyses, editorials, article comments, and literature reviews. Case reports were also excluded. Studies with patients <18 years old, critically ill patients, or Helicobacter pylori treatment were excluded.


Review of methodological quality

The case-control studies were appraised using the Joanna Briggs Institute (JBI) critical appraisal checklist for case-control studies. Both cohort and longitudinal studies were appraised using the JBI critical appraisal checklist for cohort studies, due to their methodologic similarity. [35]


Data Extraction

Each article included in the review was summarized in a table for study characteristics. The extracted attributes include the author’s name, publication year, study design, study region (country), number of participants, age, sex, study duration, and the factors confounded during the calculation of the effect size. There was also a separate table where the reported effect sizes were summarized according to the duration of PPI use.


Statistical Analysis

The statistical analysis was done using the STATA software where the reported effect sizes across studies were pooled to give an overall effect size estimate. This approach was borrowed from Lambert et al. [33] The analysis used the common-effect inverse-variance model. The I2 statistic was used to determine heterogeneity and a p-value<0.05 was held as the significance threshold.



Search Results

The initial search identified 760 articles from databases. 479 articles from Scopus, 56 from ScienceDirect, 125 from PubMed, and 100 from Google Scholar. 26 duplicates were removed. During the title and abstract screening, 698 articles were excluded following the eligibility criteria, and the remaining 36 articles were subjected to a full-text review. 21 of these articles were excluded because they did not fully satisfy the inclusion criteria. 15 final studies were included in the systematic review and 9 in the meta-analysis. The reasons for exclusion are shown in Figure 1. PRISMA flowchart was produced using PRISMA 2020?compliant flow diagrams. [36]

Figure 1: PRISMA flowchart showing the study selection process


Results of quality appraisal

3 case-control studies scored ‘fair’ and 7 scored ‘good’ on overall quality. This is shown in Table 5 in the appendix. 2 of the cohort studies scored ‘fair’ and 2 scored ‘good’ on overall quality. The longitudinal study scored ‘good’ on overall quality. This is shown in Table 6 in the appendix.

Table 2


Characteristics of included studies: A summary

This paper included 15 studies, 10 nested case-control studies, 4 cohort studies, with 2 of them including self-controlled case series, and 1 longitudinal study. The mean age of inclusion across the studies varied, with most of them including participants aged ≥ 60 years.

In the review, as seen in Table 2, only the case-control studies reported participants as CAP cases versus healthy controls. In total, this review included 108, 176 cases and 1,248,785 healthy controls in the case-controlled studies. The participants in the cohort and longitudinal studies were reported as PPI-exposed versus non-exposed controls. In regard to this, this review included 622,920 PPI-exposed participants and 4,376,684 non-exposed controls, among the cohort and longitudinal studies.


Results of included studies: A summary

From nested case-control studies

In Dublin et al., [37] the prevalence of PPI use was 12% for CAP cases and 7% for controls. Analysis showed that the odds of CAP incidence was OR=1.13 (95%CI 0.88-1.44) for current PPI use compared to non-users. In Evers et al., [38] the AOR was 2.21 (95% CI, 1.66–2.94) for CAP in current PPI users compared with non-users. García Rodríguez et al., [39] used a relative risk analysis and reported CAP risk of RR (Relative risk) = 1.16 (95% CI, 1.03–1.31) for current PPI users. The adjusted odds ratio (AOR) of PPI use for CAP was 1.18 [95% CI = 0.80 - 1.74] in Gau  et al. [40] Gulmez [41] reported an AOR of 1.5 (95% CI, 1.3-1.7), associating the current use of PPIs with CAP. In Hermos et al.,[42] current PPI use was associated with CAP with an AOR = 1.29 (95%, 1.15–1.45). In Laheij [43] the adjusted relative risk for CAP among current PPI users was ARR= 1.89 (95% CI, 1.36-2.62) compared to past users (people who had used PPIs before the study period and stopped during the study period). From the analysis in this study, the adjusted attributable risk percentage for PPI use was 42%. This means that PPI use was responsible for 42% of CAP cases or 1.05 CAP per 100 person-years. Considering that the study period was 0.42 years, this meant that 1 case of pneumonia per 226 patients (exposed and unexposed) was attributable to PPIs. Current PPI use in Meijvis et al., [44] was associated with an AOR=1.6 (95% CI 1.2–2.2) for CAP incidence. Myles et al., [16] associated current PPI use with an increased risk of pneumonia (adjusted OR=1.55, 95% CI 1.38–1.77). Sarkar et al., [45] was the only case-control study that found no association between current PPI use and CAP incidence. The AOR for CAP incidence was 1.02 (95% CI, 0.97 to 1.08).


From cohort studies

Using a cohort study design, Filion et al. [46] and Roughead et al. [47] had their participants either as PPI-exposed or PPI-non-exposed. The risk for hospitalized CAP in Filion et al. [46] was AOR=1.05 (95% CI, 0.89 to 1.25). Roughead et al. [47] found that there was an increased risk of hospitalization for CAP for those exposed to PPIs compared with the unexposed group. The reported Rate Ratio (RR) was 1.16 (95% CI, 1.11–1.22).


From cohort studies including a self-controlled case series analysis

Othman et al. [48] and Ramsay et al. [49] used the same study methodology. Participants were analyzed using a cohort study design and a self-controlled case series design. When using the cohort study design, the incidence of CAP among the PPI users was compared to non-PPI users. When using the self-controlled case series, only the group of PPI users was used to perform analysis. The analysis would compare the incidence rate of CAP during the exposure period to the incidence rate during the baseline period (no-PPI use).

By using a cohort study design, the adjusted Cox regression in Othman et al. [48] showed an increased risk of CAP at Hazard Ratio (HR) =1.67 (95% CI, 1.55 to 1.79) for PPI users compared to non-users. Ramsay et al. [49] also showed an increased risk for CAP hospitalization due to PPI use, with the highest risk seen during the first 7 days of treatment. The Adjusted rate ratio (ARR) for CAP in 0-7 days of PPI use was 3.24 (95%CI, 2.50 – 4.19).

The self-controlled case-series analysis by Ramsay et al., [49] showed that the risk for CAP in the first 7 days of PPI therapy was Adjusted rate ratio (ARR) = 3.07 (95%CI, 2.69 – 3.50). From prior event rate ratio analysis by Othman et al., [48] the rate of pneumonia for the exposed patients was similar before a PPI prescription (62.1 per 1000 person-year) to the rate after a PPI prescription (61.4 per 1000 person-year). This would translate to an incidence rate ratio of 1.19 (95 CI, 1.14 to 1.25) in the 30 days after a PPI prescription, with a higher rate ratio of 1.92 (95 CI, 1.84 to 2.00) for the 30 days before a PPI prescription.


From a longitudinal study

Zirk-Sadowski et al. [50] was the only study that explicitly looked at long-term PPI use. In the study, the second year of PPI therapy was associated with an increased risk of incident pneumonia with an adjusted hazard ratio (AHR) = 1.82, (95% CI, 1.27–2.54).

Table 3

Please click here to view complete article