Gastric acid is required to cleave vitamin B12 from ingested dietary proteins, and it is produced by the same cells that produce intrinsic factor, which is required for vitamin B12 absorption. Medicines that suppress the production of gastric acid may therefore lead to vitamin B12 malabsorption. Data so far in this area are however limited, with conflicting results.
The purpose of the current case-control study was to evaluate the relationship between use of proton pump inhibitors (PPIs) or histamine 2 receptor antagonists (H2RAs) and the risk of vitamin B12 deficiency in a large community-based population (the Kaiser-Permanente Northern California [KPNC] healthcare system). Cases were adults with a diagnosis of vitamin B12 deficiency made between 1997 and 2011; each case was matched to up to 10 controls (with no vitamin B12 deficiency diagnosis) according to sex, region, race/ethnicity, year of birth and duration of KPNC membership.
The exposure of interest was medication exposure (calculated as ‘days supplied’, based on knowledge of the number of pills dispensed and the daily dose) for PPIs and H2RAs. Exposed patients were those who received at least 2 years of supply prior to the index date (for H2RAs, the criteria for exposure also included a lack of current or prior prescription for a PPI). Unexposed patients were those who did not have current or previous prescriptions for any of these medicines. When evaluating the association, a number of confounders were considered, including conditions associated with vitamin B deficiency (e.g. dementia) and extent of healthcare utilisation.
The final analysis included 25,956 cases (12% exposed to PPIs and 4.2% H2RAs) and 184,199 controls (7.2% exposed to PPIs and 3.2% H2RAs). Compared to non-users, a new diagnosis of vitamin B12 deficiency was more common in patients receiving a ≥2-year supply of both PPIs (OR 1.65; 95% CI 1.58-1.73) and H2RAs (OR 1.25; 95% CI 1.17-1.34). The association appeared to be dose-related. Using a baseline population risk of 2.3% for people aged >50 years, a number needed to harm of 67 was calculated for 2 or more years of PPI use.
Despite a number of improvements over previous studies in this area and findings to suggest a possible causal association, there are several limitations highlighted by the authors, including the possibility of ascertainment bias (patients receiving PPIs and H2RAs may be more likely to be screened for vitamin B12 deficiency), misclassification of exposure status, and residual confounding. The authors say that their findings should inform discussions of the known benefits and possible risks of these medicines, and they recommend that clinicians exercise appropriate vigilance when prescribing them, using the lowest effective dose.