LAUSR

Background Serum chloride was recently found to be associated with prognosis of heart failure in western countries. However, the evidence was scarce in Asia. We aimed to investigated the relationship between serum chloride and clinical outcomes in a Chinese cohort with hospitalized heart failure. Methods We retrospectively analyzed the data from PhysioNet, involving 1996 patients who were admitted with heart failure between December 2016 and June 2019. Outcome was a composite endpoint of all-cause death or rehospitalization at 3 months. Results The incidence of the composite endpoint was 26.8% (535/1,996); it was 32.2% (213/662), 25.0% (165/661), and 23.3% (157/673) by chloride tertiles (from the lowest to the highest), respectively. The serum chloride at admission was independently and inversely associated with the composite endpoint risk (hazard ratio: 0.967; 95% confidence interval: 0.939 to 0.996; p = 0.026) in contrast to sodium, which was no longer significant (p > 0.05) after multivariable adjustment. Pearson correlation between serum chloride and sodium was 0.747 (p < 0.001). However, an increased AUC was not observed by adding sodium to model composed of age, sex, NYHA class, diabetes, log BNP and chloride (0.620 vs. 0.612, p = 0.132). Subgroup analysis showed the presence or absence of hyponatremia did not affect the association between chloride and composite endpoint risk. Conclusions Low serum chloride at admission was associated with poor outcomes in Chinese hospitalized patients with heart failure. These findings warrant future studies for tackling the potential pathophysiological mechanisms and correction methods of hypochloremia in heart failure.