Fluctuation of pre-hemodialysis serum sodium

Abstract

Low pre-hemodialysis (pre-HD) serum sodium or hyponatremia is associated with higher mortality. Pre-HD serum sodium can be more stable over time with low fluctuation compared to other serum parameters. We examined variation of pre-HD serum sodium in 24 months and after this point examined all-cause mortality in a cohort of 261 patients followed-up for 48.8 (standard deviation (SD)=19.1) months. 6,221 determinations of pre-HD serum sodium were made and corrected for glucose concentrations. Serum sodium was measured pre-HD monthly, and the variability was calculated using the coefficient of variation (CV). The mean age was of 60±14.1 years, 60.9% were men, 48% had diabetes mellitus, and diabetic nephropathy was the most frequent cause of end-stage renal disease. Median CV of sodium in 24 months was 1.7% with a mean of 1.78% (95% CI 1.73-1.83). Patients with CV > 1.7% had a higher mortality (53 patients a 36.8%) compared to CV < 1.7% (22 patients a 18.8%) (p=0.002). In Kaplan-Meier analysis, patients with CV > 1.7% had significantly worse overall survival (log rank=6.395, p=0.011). We also stratified the sample in serum sodium tertiles (< 138mEq/L; 138-140 mEq/L; > 140 mEq/L) and made a Kaplan-Meier analysis which showed persistent worse survival outcomes in patients with CV > 1.7% (log rank Mantel-Cox 7.64; p=0.006). Cox regression multivariate model showed that CV of sodium was significantly associated with overall mortality after adjusting for confounder variables (hazard ratio 2.16, 95% CI 1.37-3.41; p=0.001). Variation of pre-HD serum sodium in 2 years is less than a 2%. With the limitations of our study, a higher variability of pre-HD serum sodium in 2 years of treatment (CV > 1.7%) is associated with increased mortality. .