Association between serum potassium variability during hospitalization and clinical outcomes in patients with heart failure

Abstract

Background: Serum potassium levels exhibit a U-shaped relationship with prognosis in patients with heart failure, but the clinical significance of in-hospital potassium variability remains poorly understood.

Methods: The cohort study analyzed 1280 patients hospitalized for heart failure between 2015 and 2022 (mean age 70.4 ± 14.7 years, 57.2 % male). Potassium variability was assessed using the standard deviation (SD) of in-hospital potassium measurements. Cardiac events (cardiac death, worsening heart failure, acute coronary syndrome) and all-cause mortality were recorded over a median follow-up of 806 days.

Results: Among the cohort, 91 % showed to maintain normokalemia at both admission and discharge, which correlated with reduced risks of cardiac events as well as all-cause deaths, compared to those with hypokalemia or hyperkalemia at either admission or discharge (log rank, P < 0.01 for both). Potassium variability, measured by SD, was analyzed in 1165 patients with normokalemia at both admission and discharge. Multivariate regression analysis identified NYHA class III/IV, B-type natriuretic peptide, and estimated glomerular filtration rate as independent factors influencing potassium SD values. When stratified into quartiles based on potassium SD, patients in the higher quartiles of potassium SD had significantly lower event-free survival rates (log rank, P = 0.02 and < 0.01). Higher potassium SD independently predicted adverse outcomes (hazard ratios, 1.49 and 1.65; P = 0.04 and 0.01).

Conclusion: Greater serum potassium variability during hospitalization was associated with long-term adverse outcomes in heart failure patients, even in those with normal potassium ranges at both admission and discharge.

Keywords: Biomarker; Heart failure; Potassium; Prognosis; Variability.