THE EFFECT OF SERUM SODIUM ON SURVIVAL IN PATIENTS TREATED BY PERITONEAL DIALYSIS IN THE UNITED KINGDOM

Background: Studies in hemodialysis patients suggest that hyponatremia is associated with increased mortality. However, results from peritoneal dialysis (PD) patients are discordant. We wished to establish whether there was an association between serum sodium and mortality risk in PD patients. Methods: We analyzed 3,108 PD patients enrolled at day 90 of renal replacement therapy (RRT) into the UK Renal Registry (UKRR) data base with available serum sodium measurements (in 3 groups: <= 137, 138 - 140, >= 141 mmol/L) who were then followed up until death or the censoring date (31 December 2012). Analysis used Cox-regression with adjustment for age, sex, year of starting RRT, primary renal disease, serum albumin, smoking, and comorbidities. Results: Unadjusted mortality rates were 118.6/1,000 personyears (py), 83.4/1,000 py, and 83.5/1,000 py for the lowest, middle, and highest serum sodium tertiles, respectively. After adjustment for covariates, patients in the lowest serum sodium group had almost 50% increased risk of dying compared with those with the highest serum sodium (hazard ratio [HR] 1.49, confidence interval [CI]: 1.28 - 1.74), with a graded association between serum sodium and mortality. The association of serum sodium with mortality varied by age (p interaction < 0.001), and whilst this association attenuated after adjustment for confounding variables in the older age groups (55 - 64, and > 65 years), it remained in the younger age group of 18 - 54 years (HR 2.24 [1.36-3.70] in the lowest compared with the highest sodium tertile). Conclusions: Lower serum sodium concentrations at the start of RRT in PD patients are associated with increased risk of mortality. Whilst this association may well be due to confounding in the older age groups, the persistent strong association between hyponatremia and mortality in the younger age group after adjustment for the available confounders suggests that prospective studies are required to assess whether active intervention to maintain serum sodium changes outcomes.