Objective: to examine whether repeated measures of serum sodium are associated with mortality and hospitalization in a large national cohort of patients with heart failure with preserved ejection fraction (HFpEF).
Methods: We identified subjects with HFpEF using a validated algorithm in the national Veterans Affairs electronic health records from 2002 to 2012 with all recorded ejection fractions greater than 50% and signs, symptoms, and treatment of heart failure. Repeated serum sodium measurements were extracted from the date of diagnosis until censoring (end of the study period or death). In a multivariable Cox proportional hazards model, we examined the association of serum sodium over time with all-cause mortality. We then used generalized estimating equation-based negative binomial regression to compute the incidence density ratios (IDRs) per year to examine days hospitalized for heart failure and for all causes.
Results: In total, we identified 56,489 subjects with 865,124 recorded serum sodium measurements. Mean age was 71±11 years. After a median follow-up of 3.6 years (IQR: 1.7-6.4), 17,269 deaths occurred. In a multivariable model, hazard ratios for long-term mortality were 2.09 (95% CI: 2.02-2.17) for the sNA 115-134.99 category; 1.29 (95% CI: 1.25-1.33) for the sNA 135-137.99 category; and 1.66 (95% CI: 1.60-1.74) for the sNA 144-166.00 category compared to the 138-140.99 category (ref), p non-linear trend <0.0001. There were a total of 1,950,224 days of all-cause hospitalization and 112,500 days of heart-failure hospitalization. The IDRs for the lowest sNA group were 1.85 (95% CI: 1.74-1.96) for all-cause hospitalization and 2.17 (95% CI: 1.73-2.74) for heart-failure hospitalization compared to the 138-140.99 sNA group (Table 1).
Conclusions: Our findings suggest that low serum sodium over time is associated with increased risk of all-cause mortality and greater number of days of all-cause and heart failure hospitalization in patients with HFpEF.