The impact of chronic pre-dialysis hyponatremia on clinical outcomes in maintenance hemodialysis patients

Objective Chronic pre-dialysis hyponatremia is not rare in maintenance hemodialysis (MHD) patients. However, the association between chronic pre-dialysis hyponatremia and mortality is uncertain due to multiple potential confounders such as hyperglycemia, fluid overload, and malnutrition. This study aimed to more comprehensively evaluate the association between chronic pre-dialysis hyponatremia and clinical outcomes in MHD patients. Methods We analyzed the data of 194 MHD patients with regular real-time measurements of pre-dialysis serum sodium from July 2015 to March 2021. Hyponatremia was defined as SNa ≤ 135 mmol/L and normonatremia as SNa > 135 mmol/L and < 145 mmol/L. We evaluated the association of baseline pre-dialysis serum sodium (SNa) and time-averaged SNa (TASNa) levels with all-cause mortality or new major adverse cardiovascular events (MACE) in MHD patients. Furthermore, the SNa levels were glucose, serum albumin, and fluid overload adjusted. The associations between SNa levels and all-cause mortality or new MACE were analyzed using time-varying Cox regression models. Results Among the total of 194 patients, 24 patients died and 45 new MACE occurred during a mean 35.2-month follow-up period. The baseline pre-dialysis SNa level was 137.1 ± 2.8 mmol/L (127–144 mmol/L). Kaplan–Meier survival analysis showed that there were no significant differences in all-cause mortality or new MACE between hyponatremia and normonatremia groups according to baseline pre-dialysis SNa or glucose-corrected SNa (gcSNa). The mean values of both TASNa and time-averaged glucose-corrected SNa (TAgcSNa) were 136.9 ± 2.4 mmol/L and 138.3 ± 2.0 mmol/L, respectively. Kaplan–Meier survival analysis showed that patients with pre-dialysis hyponatremia had higher all-cause mortality or new MACE compared with normonatremia patients whether grouped on TASNa or TAgcSNa. Cox models showed an increased risk of all‐cause mortality and new MACE in MHD patients with pre-dialysis hyponatremia based on TASNa or TAgcSNa. Even after full adjustment including time-dependent age and dialysis vintage, gender, diabetes, time-averaged weight gain (TAWG), and serum albumin, patients with pre-dialysis hyponatremia based on TASNa (HR 2.89; 95% CI 1.18–7.04; model 3) or TAgcSNa (HR 5.03; 95% CI 1.87–13.57; model 3) had approximately twofold or fourfold greater risk of all-cause mortality, respectively, compared with those with normonatremia. The risk of new MACE was also significantly elevated in patients with pre-dialysis hyponatremia based on TASNa (HR 3.86; 95% CI 2.13–7.01; model 1) or TAgcSNa (HR 2.43; 95% CI 1.14–5.15; model 1). After adjustment for time-dependent age and dialysis vintage, gender, diabetes, TAWG, and serum albumin, patients with pre-dialysis hyponatremia based on TASNa (HR 2.33; 95% CI 1.16–4.68; model 3) had a higher risk of new MACE compared with those with normonatremia. Conclusions Pre-dialysis time-averaged hyponatremia is independently associated with increased risks of all-cause mortality or new MACE in MHD patients. The baseline SNa level is not a predictor of clinical outcomes due to its variation over time. Hyperglycemia, fluid overload, and malnutrition do not have a significant impact on the risk association between chronic hyponatremia and all-cause mortality or new MACE in MHD patients.