Serum Sodium Modifies the Association of Systolic Blood Pressure with Mortality in Peritoneal Dialysis Patients

Introduction: High serum sodium is associated with increased blood pressure (BP) in dialysis patients, which is a risk factor for cardiovascular (CV) disease. However, the interaction between serum sodium and BP and their association with clinical outcomes in peritoneal dialysis (PD) patients is uncertain. Methods: We analyzed a retrospective cohort of 1,656 incident PD patients from January 2006 to December 2013, who were followed up until December 2018. Cox proportional hazards regression models were used to evaluate the association of serum sodium and BP with all-cause and CV mortality. A priori interaction between serum sodium and systolic BP (SBP) was explored, and a subgroup analysis was performed by stratifying SBP into the following 3 groups: <110, 110-130, and >130 mm Hg. Results: Mean baseline serum sodium was 140.2 +/- 3.6 mmol/L, mean SBP was 137 +/- 20 mm Hg, and diastolic BP was 85 +/- 14 mm Hg. During a median (range) follow-up time of 46.5 (2.6-154.3) months, 507 patients died, 252 of whom died due to CV disease. SBP did not predict all-cause and CV mortality when BP was assessed as a continuous variable. However, SBP >130 or <110 mm Hg was associated with higher risk of all-cause and CV mortality compared with SBP of 110-130 mm Hg. There was a significant interaction between baseline serum sodium and SBP for all-cause mortality (p for interaction = 0.016). In subgroup analysis, among those with SBP >130 mm Hg, the risk of all-cause mortality was elevated in those with serum sodium >= 140 mmol/L (adjusted hazard ratio [aHR] 1.45 [95% confidence interval (CI): 1.07-1.98]), but not for those with serum sodium <140 mmol/L (aHR 1.27 [95% CI: 0.89-1.82]). Conversely, among those with SBP <110 mm Hg, those with serum sodium <140 mmol/L had an elevated risk of mortality (aHR 1.99 [95% CI: 1.31-3.02]), but not those with serum sodium >= 140 mmol/L (aHR 1.15 [95% CI: 0.74-1.79]) (p for interaction = 0.028). Conclusion: The association of BP with mortality was modified by serum sodium levels in PD patients. Further studies are needed to evaluate whether individualized BP control based on serum sodium levels contributes to improve patient outcomes.