Association of early serum potassium with all-cause mortality in adult maintenance hemodialysis patients / 中华肾脏病杂志

ABSTRACT

Objective:To investigate the association of early serum potassium level with all-cause mortality in adult maintenance hemodialysis (MHD) patients.Methods:It was a retrospective cohort study. The data of patients newly entered MHD in the registration system of Zhejiang province dialysis quality control center from January 1, 2010 to December 31, 2019 were collected. Follow-up was conducted until December 31, 2020. The average value of predialysis serum potassium within the first 3 months starting hemodialysis was defined as early serum potassium, and patients were divided into 6 groups according to their early serum potassium levels. Death within 1 year of MHD patients was defined as short-term death. Kaplan-Meier method was used to compare the long-term and short-term survival rates of the six groups. Cox regression model was used to analyze the association of different serum potassium levels with the short-term all-cause mortality of adult MHD patients.Results:A total of 27 362 patients aged (61.2±14.4) years old were included, including 16 775 males (61.3%), 1 303 patients (4.8%) with hypokalemia (serum potassium<3.5 mmol/L) and 10 034 patients (36.7%) with hyperkalemia (serum potassium≥5.0 mmol/L). Among them, there were 5 145 patients (18.8%) with serum potassium≥5.5 mmol/L. According to the early serum potassium levels, the patients were divided into group 1 (serum potassium<3.5 mmol/L), group 2 (3.5≤ serum potassium<4.0 mmol/L), group 3 (4.0≤serum potassium<4.5 mmol/L), group 4 (4.5≤serum potassium<5.0 mmol/L), group 5 (5.0≤serum potassium<5.5 mmol/L) and group 6 (serum potassium≥ 5.5 mmol/L), respectively. Until the end of follow-up, the follow-up time was (40.7±27.8) months and 5 400 patients died. Cardiovascular and cerebrovascular diseases [1 551 cases (28.7%)] and infections [366 cases (6.8%)] were the main causes of death. Kaplan-Meier survival analysis showed that the long-term and short-term cumulative survival rates in the serum potassium<3.5 mmol/L group were the lowest among the 6 groups (Log-rank test, χ2=119.0, P<0.001; χ2=74.6, P<0.001, respectively). Multivariate Cox regression analysis showed that early serum potassium<3.5 mmol/L was an independent influencing factor for short-term all-cause death in MHD patients (with 4.5≤serum potassium<5.0 mmol/L as reference, HR=1.54, 95% CI 1.26-1.89, P<0.001). In the subgroup of age≥65 years, multivariate Cox regression model showed that serum potassium<4.5 mmol/L was independently associated with short-term death in MHD patients (with 4.5≤ serum potassium< 5.0 mmol/L as reference, serum potassium<3.5 mmol/L, HR=2.16, 95% CI 1.69-2.75, P<0.001; 3.5≤serum potassium<4.0 mmol/L, HR=1.40, 95% CI 1.14-1.72, P=0.001; 4.0≤serum potassium< 4.5 mmol/L, HR=1.46, 95% CI 1.21-1.75, P<0.001), while in the subgroup of age<65 years, serum potassium level was not significantly associated with short-term mortality risk in MHD patients. The early serum potassium level was associated with the risk of short-term all-cause death in a "U" shape, and both low and high potassium levels increased the risk of short-term all-cause death. The optimal early blood potassium level was about 4.75 mmol/L. Conclusions:The prevalence of hypokalemia at early stage of dialysis in adult MHD patients is about 4.8%. There is a U-shaped association between early serum potassium level and short-term (1 year) all-cause mortality risk, and early serum potassium<3.5 mmol/L is an independent risk factor for long-term and short-term all-cause mortality in MHD patients.