Proton Pump Inhibitors and Risk of Hyponatremia.

Abstract

Background: Proton pump inhibitors (PPIs) have been linked to hyponatremia, but evidence based on laboratory-confirmed sodium measurements remains limited.

Methods: We conducted a retrospective population-based cohort study using the Stockholm Sodium Cohort, linking laboratory sodium measurements with nationwide health registers. To reduce confounding by indication, adults initiating treatment with PPIs were compared to adults initiating treatment with histamine-2 receptor antagonists. Outcomes were defined using serum sodium concentrations. Inverse probability of treatment weighting (IPTW) based on propensity scores was used to balance baseline covariates between groups. IPTW-weighted Cox proportional hazards models estimated hazard ratios (HRs) for profound hyponatremia (<125 mmol/L) within 90 days. Secondary analyses examined thresholds of Na<130 and <135 mmol/L.

Results: A total of 492,994 individuals were included (451,676 using PPIs, 41,318 using histamine-2 receptor antagonists). During follow-up, 1,104 PPI users and 27 histamine-2 receptor antagonists users developed profound hyponatremia. After IPTW, PPI use was associated with higher risk of profound hyponatremia (HR 2.57; 95% CI 1.74-3.81). Hyponatremia defined as Na<130 mmol/L occurred in 2,088 PPI users and 76 histamine-2 receptor antagonists' users (HR 1.72; 95% CI 1.36-2.18). Na<135 mmol/L developed in 11,226 PPI users and 615 histamine-2 receptor antagonists' users (HR 1.43; 95% CI 1.31-1.56). Events clustered early after treatment initiation. The IPTW-weighted 90-day absolute risk of profound hyponatremia was 0.24% among PPI initiators and 0.09% among histamine-2 receptor antagonists' initiators (number needed to harm, 680).

Conclusions: PPI-treatment was associated with a higher risk of hyponatremia compared to treatment with histamine-2 receptor antagonists. This supports a strong independent association between use of PPIs and profound hyponatremia.