Factors associated with one-year mortality in patients with hyponatremia hospitalized in a cardiology department: An exploratory retrospective cohort study

BACKGROUNDHyponatremia is a common electrolyte disorder in hospitalized cardiac patients and is associated with increased mortality.However, data on factors associated with one-year mortality in patients with hyponatremia in real-world clinical practice are limited, and the temporal relationship between in-hospital predictors and long-term outcomes remains uncertain. AIMTo explore factors associated with one-year mortality in cardiac inpatients with hyponatremia and assess the robustness of these associations in sensitivity analyses. METHODSThis retrospective single-center cohort study included 119 patients (mean age 71.2 ± 12.7 years, 48.7% female) hospitalized in a cardiology department with hypotonic hyponatremia.Patients with acute coronary syndrome and/or myocardial infarction were excluded.The primary endpoint was all-cause death within 1 year, ascertained through the regional electronic death registry.Cox regression was used.Sensitivity analyses excluding in-hospital deaths and Firth's penalized Cox regression were performed. RESULTSThe one-year mortality rate was 44.5% (53/119).In the primary multivariable Cox model, stage II chronic heart failure (CHF) [hazard ratio (HR) 8.86; 95% confidence interval (CI):1.81-43.39;P = 0.007], minimum sodium (HR 0.94; 95%CI: 0.89-1.00;P = 0.040), and peak creatine kinase-MB (CK-MB) (HR 1.02; 95%CI: 1.00-1.04;P = 0.002) were associated with one-year mortality.However, in a sensitivity analysis excluding in-hospital deaths, these associations were no longer statistically significant.In Firth's penalized model, CK-MB 4 / 35 (HR 1.02; 95%CI: 1.01-1.04;P = 0.007) and stage II CHF (HR 6.60; 95%CI: 1.67-33.37;P = 0.007) remained significant, whereas minimum sodium did not (P = 0.117).Non-linearity was not detected for sodium (P = 0.459) or CK-MB (P = 0.182). CONCLUSIONIn this exploratory analysis, CK-MB and stage II CHF were more robustly associated with one-year mortality than hyponatremia.The sodium association was not confirmed in sensitivity analyses.These findings are hypothesis-generating and require prospective validation.