(P-169) The Association Between Proton Pump Inhibitors and Increased Risk of Hyponatremia - A Cohort Study

Background: Prescribing and dispensing of Proton Pump Inhibitors (PPIs), especially in the elderly, have been profoundly increased in the last years. However, several safety concerns were raised regarding their potential adverse events. In recent years, a few studies and case reports suggested a possible association between PPIs and hyponatremia.

Objectives: To determine the hazard ratio of hyponatremia among PPI's users compared to controls.

Methods: This retrospective cohort study design. We used databases from the Clalit Research Secure Data Sharing Platform in Israel's central district, comprising data from the community and hospital settings. The study population included all adult populations above 18 in 2000-2018 with no further exclusion criteria. The exposed group comprised patients who purchased PPI for the first time, set as the index date. Similarly, the control group included patients who purchased H2-receptor blockers (H2RB). Exposure duration was determined by the total amount dispensed during the study period, considering the daily dosages. The participants were followed for 90 days after treatment initiation for new users’ analysis, whereas a maximum follow-up of 2 years for chronic users’ analysis after dispensing to monitor hyponatremia events. Hazard ratios (HRs) for acute exposure were calculated for 90 first days with a 95% confidence interval (95% CI) using Cox regression for each group. Age, sex, diuretics exposure, and hyponatremia-associated comorbidities were adjusted in time-varying methodology.

Results: 521,344 patients registered to Clalit health maintenance organizations met the inclusion criteria, of which 337,531 (64.7%) patients were exposed to PPIs, and 183,813 (35.3%) patients were exposed to H2B in the new users' analysis. Adjusted HR in Gross Model indicated a mild decrease in hyponatremia HR in PPIs treatment vs. control treatment (adjusted-HR = 0.915, P-value = 0.003, respectively). Evaluation of severity distribution of hyponatremic events does not indicate that PPIs induce more severe hyponatremia than H2RBs. In chronic exposure analysis, adjusted HR was significantly lower for PPI’s in comparison to H2B’s in two examined periods (adjusted-HR180-365 days = 0.842, p-value = 0.001, adjusted-HR365-730 days = 0.822, p-value < 0.001).

Conclusions: A conservative approach suggests that new and chronic exposure to PPI’s are not associated with hyponatremia compared to H2Bs users.