Do We Need Our Gastric Acid? Risk of Bacterial Gastroenteritis With Proton Pump Inhibitors

Article Outline

• Comment
• Copyright

Garcia Rodriguez LA, Ruigomez A, Panes J. (Centro Espanol de Investigacion Farmacoepidemiologica, Madrid Spain; and Servei de Gastroenterologia, Hospital Clinic de Barcelona, Institut d'Investigacions Biomediques August Pi I Sunyer, Centro de Investigacion Biomedica en Red de Enfermedades Hepaticas y Digestivas, Barcelona, Spain). Use of acid-suppressing drugs and the risk of bacterial gastroenteritis. Clin Gastroenterol Hepatol 2007;5:1418–1423.

Previous studies have suggested that the use of acid-suppressing medications may increase the risk of bacterial gastroenteritis. To examine this possible association, the authors used the United Kingdom's General Practice Research Database, a population-based database with clinical information (including diagnoses and records of all prescriptions) from several million patients.

A nested case-control approach was used. The study population consisted of persons 20–74 years old with no history of gastroenteritis or inflammatory bowel disease. Cases were defined as having symptoms consistent with gastroenteritis and a stool culture positive for a specific bacteria (Salmonella, Campylobacter, Shigella, or C Difficile). Controls were selected randomly from the same study population, frequency-matched to cases by age, gender, and calendar year. Exposure to acid suppression was determined by review of prescriptions and categorized by the length of prescribed drug therapy: current (most recent prescription lasting until 0–6 days before the episode of gastroenteritis), recent (prescription ending 7–90 days before), past (91–365 days before), and nonuse (no prescription issued within the past year). Among current users of proton pump inhibitors (PPI) or H2 receptor antagonists (H2RA), the effect of duration, dose, and treatment indication were studied. Daily dose was categorized as low, medium, or high (using the standard prescribing dose as the cut off value for medium). Relative risks (RR) of gastroenteritis in users of PPI or H2RA were calculated. Estimates were adjusted for age, gender, calendar year, and number of visits to the general practitioner within the year prior to the episode of gastroenteritis.

The authors identified 6,414 cases (Campylobacter, n = 4,124; Salmonella, n = 1,885; Shigella, n = 312; C Difficile, n = 31) and 50,000 controls. Compared with nonusers, current PPI use was associated with an increased risk of gastroenteritis (2.9; 95% confidence interval [CI], 2.5–3.5). The risk was also increased, to a lesser degree, with recent use (RR, 1.45; 95% CI, 1.10–1.95). The risk for past users was similar to nonusers (RR, 1.1; 95% CI, 0.8–1.4). PPI dose was associated with a progressive increase in risk: low-dose, RR 2.69 (95% CI, 1.85–3.91), medium-dose RR, 2.67 (95% CI, 2.16–3.31), and high-dose RR, 5.02 (95% CI, 2.71–9.31). The risk was greater in PPI users administered the high dose in divided doses (BID; RR, 5.3; 95% CI, 2.8–10.3) than in those administered the high dose on a once daily basis (RR, 3.2; 95% CI, 0.5–19.4). H2RA use was not associated with an increased risk.

The authors conclude that acid suppression with PPI is associated with an increased risk of Salmonella and Campylobacter gastroenteritis.

Back to Article Outline

Comment 

PPIs are extremely effective medications for the management of acid-related diseases such as peptic ulcer disease (PUD) and gastroesophageal reflux disease (GERD). The effectiveness of PPI, the high prevalence of acid-related disorders, the expanding indications for PPI use, and the favorable side effect profile have resulted in millions of chronic PPI users. This trend has raised concerns regarding the long-term safety of these medications. Potential long-term effects of PPI-induced acid suppression include gastric carcinoid formation, development of gastric adenocarcinoma, bacterial overgrowth, and malabsorption of vitamins and minerals (eg, vitamin B₁₂, iron, and calcium). Because most bacteria cannot survive an acidic environment, gastric acid is considered the first line of defense against ingested microorganisms. Purposeful suppression of gastric acid, therefore, has also been speculated to increase the risk of bacterial gastroenteritis.

Multiple studies to assess for an association between acid-suppressing medications and enteric infections have been performed and report conflicting results. A recent systematic review was conducted to further evaluate for an association (Am J Gastro 2007;102:2047-56). Six observational studies on community acquired bacterial gastroenteritis (e.g. Campylobacter, Salmonella, and Shigella) were identified, consisting of a total of 11,280 patients (BMJ 1994;308:176; BMJ 1996;312: 414-5; Epidemiol Infec 1997;119:307-11; Epidemiology 1997; 8:571-4; Epidemiol Infect 1999;122:201-7; Epidemiol Infect 2006;134:617-26). There was a significant association between acid suppression and bacterial gastroenteritis with an OR of 2.55 (95% CI 1.53-4.26). The association was greater for PPI use than H2RA use, OR 3.33 (95% CI 1.84-6.02) and OR 2.03 (95% CI 1.05-3.92), respectively. The stronger association for PPI use than H2RA use suggests a dose-response relationship. The current study by Garcia Rodriguez seeks to better assess for a potential causal relationship between acid suppression and bacterial gastroenteritis by improving upon limitations of these previous studies. The authors sought to better establish a dose response relationship between PPI exposure and risk by looking at medication dose and duration of treatment prior to the episode of gastroenteritis.

The strengths of the study include the use of a large, population-based database (General Practice Research Database) that has been used to perform >500 published epidemiologic studies. The nested, case-control study design and the use of issued prescriptions to document exposure help to minimize recall bias and establish a temporal relationship between acid suppression and the episode of bacterial gastroenteritis. However, compliance, intermittent use, and undetected over-the-counter use could affect the true exposure in both groups.

The findings clearly support a dose response relationship between acid suppression and risk of gastroenteritis as demonstrated by (1) the risk of gastroenteritis rapidly returning to baseline with discontinuation of PPI use, (2) the risk of gastroenteritis increasing with the more potent acid suppressing PPI, but not H2RA, (3) the risk increasing with an increasing dose of PPI, and (4) the risk increasing with continuous rather than interrupted acid suppression (twice daily vs once daily dosing). The dose-dependent increased risk of gastroenteritis with gastric acid suppression is consistent with the theory that gastric acid is an important defense mechanism against enteric pathogens.

The RR of gastroenteritis in current PPI users did vary slightly with treatment indication. It is interesting to note that the RR of current PPI users treated for dyspepsia or epigastric pain was 3.87 (95% CI, 2.26–6.65) compared with 2.93 (2.36–3.63), 2.57 (1.58–4.2), and 1.6 (0.84–3.03) for GERD, PUD, and nonsteroidal anti-inflammatory drug prophylaxis, respectively. Possible explanations for the association of functional-type gastrointestinal (GI) symptoms and gastroenteritis, other than acid suppression, include decreased clearance of organisms and/or increased duration of exposure to pathogenic organisms owing to underlying dysmotility or ascertainment bias owing to a greater likelihood of patients with functional disorders to seek medical evaluation and have stool cultures ordered.

Another interesting finding was the association between gastroenteritis and the number of office visits within the year prior to gastroenteritis. The RR of gastroenteritis progressively increased in patients with 2–4, 5–9, and ≥10 visits in the year prior compared with 0–1 visit (2.26, 3.15, and 4.19, respectively). This difference in risk of gastroenteritis based on the frequency of physician visits raises the concern that there may be differences between the groups, other than acid suppression, that might influence their risk of gastroenteritis (eg, chronically ill, multiple comorbidities, greater likelihood of having stool tested). Similar to previous studies, only laboratory-confirmed cases were included. These cases are a considerable underestimate of the actual number of bacterial enteric infections because many patients do not seek medical attention and testing is not always done when patients do contact their physician. Considering only culture-positive cases might overestimate the proportion of cases exposed to PPI, if those patients are more likely to seek care. Alternatively, in line with the biologically feasible hypothesis, impaired defense owing to acid suppression may expose an individual to greater quantities of bacteria which could result in more severe symptoms, prompting medical evaluation and stool testing.

It remains difficult to determine whether PPI use is a true risk factor for bacterial gastroenteritis. The theory that gastric acid is important in eliminating ingested bacteria and that suppression of gastric acid would result in increased susceptibility to infection is reasonable. However, when gastric acid is decreased owing to ulcer surgery or aging (atrophic gastritis), most individuals seem to be little affected. The GI tract has other defense mechanisms that influence the pathogenicity of ingested bacteria and would remain intact in patients on acid suppression. Bacteria that survive passage through the stomach must then compete with the normal intestinal microbial flora and must withstand defenses such as bile salts and pancreatic enzymes. The innate immune system, cell-mediated immunity, and humoral immunity also play important roles in limiting invasion of bacteria.

Although infectious diarrhea in developed countries is generally self-limited, subgroups (immunocompromised individuals, the young, and the elderly) are susceptible to developing more severe disease. In addition to acute morbidity and mortality, long-term sequelae such as hemolytic uremic syndrome with renal failure, Guillain-Barre′ syndrome, and irritable bowel syndrome can occur. Therefore, even if the association remains unclear at his time, this study reminds the clinician that no medication is without potential adverse effects. Concerns for patient safety should guide short- and long-term use of even the most apparently benign drug. PPIs are commonly overprescribed, with many patients taking PPIs chronically for marginal indications (Ann Pharmacother 2006;40:1261–1266). Long-term use of any medication should be clinically indicated. Opportunities to minimize unnecessary exposure should be sought and could include an initial trial of H2RA therapy before PPI; an attempt to reduce the dose of PPI once GERD symptoms are controlled (Am J Gastro 2003;98:1940–1944); and an attempt to discontinue PPI use in long-term PPI users (Aliment Pharmacol Ther 2006;24:945–954).