AIMS Prior cohorts demonstrating the importance of serum chloride levels in heart failure either excluded or had partial representation of patients with heart failure with preserved ejection fraction (HFpEF). We aimed to examine the relationship between serum chloride concentration and outcomes in HFpEF. METHODS AND RESULTS We included participants from the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist Trial (TOPCAT) who met the following criteria: met inclusion by the natriuretic peptide stratum, had recorded serum chloride levels, and were from the Americas (n = 942). Multivariable Cox proportional hazards models tested the association of serum chloride with clinical outcomes, and mixed effects modelling tested the association of spironolactone or loop diuretic on serial serum chloride levels. The median serum chloride level was 102 [25th-75th percentile 100-105 mmol/L (range 84-114 mmol/L)]. After multivariable adjustment, every standard deviation decrease in serum chloride (4.05 mmol/L) was associated with ∼50% increased risk for cardiovascular death [hazard ratio (HR) 1.51, 95% confidence interval (CI) 1.11-2.06, P = 0.008] and ∼30% increased risk for all-cause death (HR 1.29, 95% CI 1.02-1.62, P = 0.04), but not with the primary composite endpoint or heart failure hospitalization (P > 0.3 for both). There were no significant interactions between spironolactone use and the serum chloride-risk relationship (P > 0.1) for each endpoint. Spironolactone was not (P = 0.33) but loop diuretic use was associated with lower serial serum chloride levels (P
