Background: Patients with heart failure (HF) are known to be stratified the risk of mortality by serum chloride level. However, since large-scaled studies to date have not included the elderly patients, it is still unclear whether serum chloride level well predicts the prognosis of the elderly patients with HF, whose clinical profiles are different from young adult patients. We aimed to test the hypothesis that hypochloremia gave different prognostic impacts in non-elderly and elderly patients with HF. Methods: This observational study included 1,326 elderly patients ( > 65 year-old) without regular hemodialysis who were hospitalized for worsening of HF and discharged alive. They were divided into two subgroups depending on serum chloride levels at admission (Low-chloride group ( < 97 mEq/L, n=244); Normal-chloride group ( > 98 mEq/L, n=1518)), referred to prior reports. Propensity scores were developed and patients in Normal-chloride group were matched with those in Low-chloride group in a 1:1 ratio. The endpoint of this study was death from any cause. Results: Among 1,326 unmatched patients, Low-chloride group included significantly more male patients, and those with higher BUN, CRP, bilirubin levels, dose of furosemide and atrial fibrillation, and prescriptions of oral anticoagulant, and lower BMI, blood pressure, hemoglobin, prescription of mineral corticoid receptor antagonist, thiazide and statins. Following propensity score matching by diverse baseline parameters, 106 pairs in both groups were included in the matched population. During 413 days of median observational period, 62 patients (28.8%) died. Kaplan-Meier analysis showed that patients in the Low-chloride group had a significantly higher mortality rate (p=0.011, Figure ). Conclusion: In this observational study, baseline hypochloremia was associated with an increased mortality in elderly patients hospitalized with heart failure.
Read full abstract