Introduction
Potassium imbalances are a common electrolyte disorder in patients with chronic kidney disease (CKD).[1-3] [4] [4-6] [5,7,8] [9]
Malnutrition is one of the predominant causes of hypokalemia in dialysis patients.[10] albumin has been verified to be positively associated with potassium in dialysis patients in our previous studies.[4,11] albumin as a mediator for consideration of nutritional status.
Methods
Patient selection
Data were collected from the national multicenter HD and PD registry databases maintained by the Department of Nephrology of the First Affiliated Hospital of Sun Yat-Sen University (https://hd.medidata.cn/ https://pd.medidata.cn/ https://links.lww.com/CM9/B462
Ethical approval
The study was conducted in accordance with the Declaration of Helsinki , and approved by the Clinical Research Ethics Committee of The First Affiliated Hospital of Sun Yat-Sen University (No. [2016]215, and all patients provided written informed consent before enrollment).
Data collection
Demographic and clinical data, including age, sex, height, weight, underlying kidney disease, and complications such as diabetes, hypertension, cerebrovascular disease, cardiovascular disease (CVD), heart failure, coronary heart disease, arrhythmia, and peripheral vascular disease, were collected. Standard laboratory variables included hemoglobin, serum potassium, carbonate, serum albumin , fasting blood uric acid, creatinine, blood calcium, phosphorus, alkaline phosphatase, blood lipid series, intact parathyroid hormone (iPTH), hypersensitive C-reactive protein, and ferritin. All blood samples were measured using commercial kits and automated analyzers. Medications included angiotensin-converting enzyme inhibitors/angiotensin receptor blockers (ACEIs/ARBs), beta-blockers, diuretics, insulin, etc. The lab measurements were one-time cross-sectional measures.
According to the 2020 Kidney Disease Improving Global Outcomes standard, hyperkalemia was defined as a serum potassium level ≥5.0 mmol/L, severe hyperkalemia was defined as a serum potassium level ≥6.5 mmol/L, and hypokalemia was defined as a serum potassium level <3.5 mmol/L.[10]
Outcomes
Patients were followed up at dialysis clinic every 3 to 6 months until April 30, 2021. The primary outcomes of this study included all-cause (AC) and cardiovascular (CV) mortalities. CV mortality was defined as death from acute myocardial infarction, ischemic heart disease, cardiomyopathy, fatal arrhythmia, cardiac arrest, congestive heart failure, cerebrovascular accident (including intracranial hemorrhage and subdural hematoma, cerebral infarction), and peripheral vascular disease.[12]
Statistical analysis
The results of continuous variables were expressed as the mean ± standard deviation for normal distributions and as the median (25th percentile and 75th percentile) for skewed distributions. Categorical variables are expressed as frequencies and percentages. Pairwise deletion (available case analysis) was used to address the missing data. Comparisons between two groups were performed by Student's t -test for continuous variables with a normal distribution and by the non-parametric Mann–Whitney test for continuous variables with a skewed distribution. Categorical data were analyzed by chi-squared tests. Restrictive cubic splines with five knots at the 5th, 35th, 50th, 65th, and, 95th percentiles were used to test the linearity of serum potassium concentrations and relationships with AC and CV mortalities. A two-line piecewise linear model was fitted by trying all possible values to approach the change point with the highest likelihood. Cox regression analysis was used to evaluate the association between hyperkalemia and AC and CV mortalities. We hypothesized that albumin might be a mediator in the relationship between potassium and mortality. Hence, we performed mediation analysis based on the Barron–Kenny method [13] albumin to potassium and mortality. IBM SPSS software (Version 22, SPSS Inc., Chicago, IL, USA) and R 4.0.2 (https://www.r-project.org/ P value <0.05 was considered statistically significant.
Results
Baseline characteristics of the study population
A total of 10,027 patients were included in this study, of whom 6605 were PD patients and 3422 were HD patients. The patient screening flow chart is shown in Figure 1 . Demographic, clinical, and laboratory values and medications are summarized in Table 1 . In the overall population, the mean age was 51.7 ± 14.8 years, 55.5% were male, and the median dialysis vintage was 13.60 (4.70, 39.70) months. A total of 15.1% (1514/10027) of the cohort had diabetes, 43.1% (4324/10027) had a history of CVD, and 78 (0.8%) had a history of arrhythmia. Baseline serum potassium was 4.30 ± 0.88 mmol/L. When serum potassium was categorized into different levels, 20.4% (2043/10027) of the patients had hyperkalemia, and 17.2 (1729/10027) had hypokalemia. Additionally, 1.5% (153/10027) had severe hyperkalemia.
Figure 1: Subjects’ inclusion flowchart. HD: Hemodialysis; PD: Peritoneal dialysis.
Table 1 -
Baseline characteristics of PD and HD patients.
Variables
Total (n = 10,027)
PD (n = 6605)
HD (n = 3422)
t /Z /χ
2
P values
Age (years)
51.7 ± 14.8
49.2 ± 13.9
56.4 ± 15.3
23.39∗
<0.001
Male, n (%)
5546 (55.3%)
3481 (52.7%)
2065 (60.3%)
113.56†
<0.001
History of diabetes, n (%)
1434 (14.3%)
757 (11.5)
677 (19.8%)
4175.24†
<0.001
History of CVD, n (%)
4324 (43.1%)
2860 (43.3%)
1464 (42.8%)
189.14†
<0.001
History of arrhythmia, n (%)
78 (0.8%)
35 (0.5%)
43 (1.3%)
7141.27†
<0.001
Dialysis duration (months)
13.60 (4.70, 39.70)
13.61 (2.33, 40.73)
13.62 (2.33, 40.73)
−9.58‡
<0.001
Serum potassium (mmol/L)
4.30 ± 0.88
4.05 ± 0.74
4.79 ± 0.91
41.75∗
<0.001
Hyperkalemia, n (%)
2043 (20.4%)
667 (10.1)
1376 (40.2%)
3520.04†
<0.001
Severe hyperkalemia (n , %)
153 (1.5)
18 (0.3)
135 (3.9)
9424.24†
<0.001
Hypokalemia, n (%)
1729 (17.2%)
1506 (22.8)
223 (6.5%)
4303.56†
<0.001
BMI (kg/m2 )
22.15 ± 3.48
22.10 ± 3.41
22.27 ± 3.66
1.61∗
0.067
Hemoglobin (g/L)
103.69 ± 22.62
105.87 ± 22.55
99.56 ± 22.19
−13.17∗
<0.001
Bicarbonate (mmol/L)
23.95 ± 4.51
25.59 ± 3.59
20.28 ± 4.19
−56.09∗
<0.001
Serum creatinine (μmol/L)
902.75 ± 325.13
917.42 ± 318.96
873.86 ± 335.12
−5.63∗
<0.001
Albumin (g/dL)36.69 ± 5.55
36.23 ± 5.57
37.63 ± 5.39
12.14∗
<0.001
Calcium (mmol/L)
2.19 ± 0.26
2.22 ± 0.25
2.14 ± 0.28
−13.98∗
<0.001
Phosphorus (mmol/L)
1.70 ± 0.59
1.60 ± 0.51
1.93 ± 0.67
25.44∗
<0.001
iPTH (pg/mL)
294.90 (152.00, 539.10)
286.00 (144.20, 526.85)
286.00 (144.20, 526.85)
−1.67‡
0.105
Alkaline phosphatase (U/L)
83.00 (64.00, 112.00)
84.00 (65.00, 116.00)
84.00 (65.00, 116.00)
−3.38‡
<0.001
Glucose (mmol/L)
5.97 ± 2.78
5.65 ± 2.31
6.67 ± 3.52
15.15∗
<0.001
Uric acid (μmol/L)
413.38 ± 103.98
402.07 ± 89.94
439.23 ± 126.78
15.67∗
<0.001
LDL-c (mmol/L)
2.61 ± 0.91
2.73 ± 0.89
2.29 ± 0.90
−21.10∗
<0.001
Cholesterol (mmol/L)
4.66 ± 1.26
4.88 ± 1.20
4.17 ± 1.25
−25.39∗
<0.001
Triglyceride (mmol/L)
1.79 ± 1.38
1.86 ± 1.48
1.64 ± 1.09
−8.14∗
<0.001
hs-CRP (mg/L)
2.75 (0.88, 7.32)
3.03 (1.10, 9.43)
3.03 (1.10, 9.43)
−7.73‡
<0.001
Ferritin (μg/L)
165.48 (65.40, 363.80)
137.17 (56.20, 287.46)
259.44 (97.53, 578.30)
−21.04‡
<0.001
Medications, n (%)
ACEI/ARB
3544 (35.3%)
2770 (41.9)
774 (22.6%)
131.64†
<0.001
β-blocker
3148 (31.4%)
2359 (35.7%)
789 (23.1%)
424.99†
<0.001
Diuretics
883 (8.8%)
708 (10.7)
175 (5.1%)
4746.63†
0.001
Insulin
517 (5.2%)
439 (6.6%)
78 (2.3%)
5897.56†
0.668
The results are expressed as means ± standard deviation for normal distributed continuous variables, median (25th percentile and 75th percentile) for skew continuous variables, or number (%) for categorical variables. Hyperkalemia was defined as serum potassium ≥5.0 mmol/L, severe hypokalemia was defined as serum potassium ≥6.5 mmol/L, and hypokalemia was defined as serum potassium <3.5 mmol/L.
∗ t value.
† χ 2
‡ Z value.ACEI/ARB: Angiotensin-converting enzyme inhibitors/angiotensin receptor blockers; BMI: Body mass index; CVD: Cardiovascular disease; HD: Hemodialysis; hs-CRP: High sensitive C-reactive protein; iPTH: Intact parathyroid hormone; LDL-c: Low-density lipoprotein cholesterol; PD: Peritoneal dialysis.
When classifying patients by dialysis modality, it was found that compared to patients with PD, patients undergoing HD were older (56.4 ± 15.3 vs . 49.2 ± 13.9, P < 0.001, respectively) and had a higher proportion of men (60.2% vs . 52.7%, P < 0.001), diabetes (19.8% vs . 11.5%, P < 0.001), and arrhythmia (5.5% vs . 0.6%). Meanwhile, HD patients had a higher level of potassium (4.79 ± 0.91 vs . 4.05 ± 0.74, P < 0.001), a higher proportion of hyperkalemia (40.2% vs . 10.1%, P < 0.001), and severe hyperkalemia (3.9% vs . 0.3%, P < 0.001) but a much lower proportion of hypokalemia (6.5% vs . 22.8%, P < 0.001). For medications, HD patients had less prescriptions of ACEIs/ARBs (52.2% vs . 41.9%, P < 0.001), β-blockers (51.8% vs . 35.5%, P < 0.001), and diuretics (13.8% vs . 10.7%, P < 0.001).
Independent and interactive associations of serum potassium and dialysis modalities with mortality
During a median follow-up period of 26.87 (14.77, 41.50) months, 1198 (11.9%) patients died. Among these deaths, 525 (43.8%) were caused by CVD, while 146 (12.2%) by infectious disease. The clinical outcomes of the study patients are shown in Table 2 . To evaluate the associations with mortality and potassium more effectively and accurately, restricted cubic splines were used to assess their linear and non-linear associations [Figure 2 ]. It was found that there were non-linear associations of serum potassium with AC and CV mortalities (P < 0.001 for both) in the overall population. Associations between potassium and mortality were U-shaped in the overall population. A marked increase in risk was observed at lower potassium, but a moderate elevation in risk was observed at higher potassium in the overall population.
Table 2 -
Outcomes of the study cohort.
Variables
n (%)
Follow up (months)
26.87 (14.77, 41.50)
Deaths
1198 (11.9%)
Cardiovascular death
525 (43.8)
Infection
146 (12.2)
Gastrointestinal hemorrhage
32 (2.7)
Tumors
36 (3.0)
Others
459 (38.3)
Kidney transplantation
511 (5.1)
Dialysis method adjusted
544 (5.4)
Note: Values are median (25–75%) or n (%).
Figure 2: Linear and non-linear correlation between serum potassium and AC and CV mortalities. Association of potassium and (A) AC mortality, (B) CV mortality in overall population; association of potassium and (C) AC mortality, (D) CV mortality in PD patients; association of potassium and (E) AC mortality, (F) CV mortality in HD patients. AC: All-cause; CV: Cardiovascular; HD: Hemodialysis; PD: Peritoneal dialysis.
When patients were categorized into PD and HD, this association was observed for PD but not HD patients. Additionally, significant interactions were detected between the potassium and dialysis modalities for AC mortality after adjusting for confounding variables (P for interaction = 0.037, Supplementary Table 2, https://links.lww.com/CM9/B424
Mediation of albumin in the associations of serum potassium and mortality at different potassium levels in the overall population
In the overall population, by trying all possible values for the changing points with the highest likelihoods in the non-linearity models in restricted cubic splines, two-line piecewise-linear models were used. The nadir for AC mortality risk was estimated from piecewise linear models to be a potassium concentration of 4.0 mmol/L. Therefore, a level of 4.0 mmol/L was chosen for a categorical split for serum potassium. Patients were then categorized into groups based on potassium concentrations overall and for PD and HD patients separately. The estimated hazard ratio (HR) for AC mortality per 1 mmol/L increase in potassium was 0.74 (95% CI: 0.59–0.94, P = 0.011) below 4 mmol/L and 0.97 (0.84–1.12, P = 0.644) above this point after adjustments for several confounders, including age, sex, history of diabetes, CVD, dialysis duration, uric acid, bicarbonate, hemoglobin, cholesterol, triglycerides, low-density lipoprotein cholesterol, phosphorus, calcium, and iPTH, in the overall population. For CV mortality, no significant associations were observed for patients either below or above this point (Table 3 ).
Table 3 -
Cox regression of serum potassium and mortality at different potassium levels in the overall population.
All-cause mortality
Cardiovascular mortality
Potassium < 4.0 mmol/L
Potassium ≥ 4.0 mmol/L
Potassium < 4.0 mmol/L
Potassium ≥ 4.0 mmol/L
Models
HR (95% CI)
P values
HR (95% CI)
P values
HR (95% CI)
P values
HR (95% CI)
P values
Model 1
0.69 (0.58, 0.82)
<0.001
1.13 (1.03, 1.24)
0.011
0.69 (0.58, 0.82)
<0.001
1.29 (1.03, 1.24)
0.011
Model 2
0.77 (0.65, 0.91)
0.003
1.06 (0.96, 1.16)
0.240
0.86 (0.66, 1.12)
0.270
0.94 (0.81, 1.09)
0.390
Model 3
0.75 (0.63, 0.90)
0.002
1.07 (0.96, 1.19)
0.219
0.78 (0.59, 1.02)
0.071
0.94 (0.82, 1.11)
0.500
Model 4
0.76 (0.64, 0.92)
0.004
1.03 (0.93, 1.14)
0.578
0.87 (0.66, 1.15)
0.325
0.89 (0.75, 1.05)
0.160
Model 5
0.74 (0.59, 0.94)
0.011
0.97 (0.84, 1.12)
0.644
0.92 (0.65, 1.31)
0.657
0.89 (0.72, 1.10)
0.280
Model 6
0.83 (0.66, 1.04)
0.106
1.01 (0.87, 1.16)
0.927
1.02 (0.72, 1.44)
0.928
0.92 (0.75, 1.14)
0.456
Model 1: unadjusted.Model 2: adjusted for age and sex.Model 3: Model 2 + DM + CVD history.Model 4: Model 3 + dialysis duration.Model 5: Model 4 + uric acid + bicarbonate + hemoglobin + cholesterol + triglyceride + LDL-c + phosphorus + calcium + iPTH.Model 6: Model 5 + albumin .CVD: Cardiovascular disease; DM: Diabetes; HR: Hazard ratio; iPTH: Intact parathyroid hormone; LDL-c: Low-density lipoprotein cholesterol.
Interestingly, it is notable that the significance of the association between potassium and mortality was attenuated when albumin was introduced into the extended adjusted Cox regression model mentioned above, suggesting that albumin substantially affects their relationship. While this attenuation may provide evidence that albumin confounds the true relationship between potassium and mortality, it could also suggest that albumin is a mediator of the effect of potassium on mortality; in particular, there are some previous studies that used potassium as a parameter of nutritional status. Therefore, a mediation analysis was conducted between potassium, albumin , and mortality. Figure 3 shows the results of an analysis that explored whether albumin was a mediator in the link between potassium and AC and CV mortalities. Significant mediations by albumin for potassium and AC mortality (Figure 3 A, P < 0.001) and CV mortality (Figure 3 B, P < 0.001) were found.
Figure 3: Analysis of albumin as a mediator of the association between potassium and (A) AC mortality and (B) CV mortality. A, pathway A1B1 for the association with potassium mediated by albumin and CV mortality P < 0.001; B, pathway A2B2: the association with potassium mediated by albumin , P < 0.001. Units of potassium and albumin are mmol/L and g/L, respectively. AC: All-cause; CV: Cardiovascular.
Associations of serum potassium and mortality at different potassium levels in the PD and HD populations
As there was a notable interaction between potassium and dialysis modality, patients were divided into PD and HD groups, and the associations between different potassium levels and mortality were further tested sequentially [Table 4 ]. After adjusting for the confounders mentioned above, the shape of the association for PD was similar to that in the overall population, and the estimated HR for AC mortality per 1 mmol/L increase in potassium was 0.71 (95% CI: 0.55–0.91, P = 0.006) below 4 mmol/L and 1.08 (95% CI: 0.88–1.31, P = 0.479) above this point. Similarly, no significant associations were observed regarding CV mortality for patients either below or above this point. In contrast, the shape of the association varied for HD patients. We did not detect a significant association with AC or CV mortality for potassium levels below or above 4.0 mmol/L.
Table 4 -
Cox regression of serum potassium and mortality at different potassium levels in the PD and HD populations.
All-cause mortality
CVD mortality
Potassium < 4.0 mmol/L
Potassium ≥ 4.0 mmol/L
Potassium < 4.0 mmol/L
Potassium ≥ 4.0 mmol/L
Models
HR (95% CI)
P values
HR (95% CI)
P values
HR (95% CI)
P values
HR (95% CI)
P values
HD
HD
Model 1
0.64 (0.46, 0.89)
0.009
0.87 (0.77, 0.99)
0.033
0.82 (0.45, 1.53)
0.541
0.80 (0.65, 1.00)
0.048
Model 2
0.72 (0.52, 1.01)
0.056
0.94 (0.83, 1.07)
0.383
0.87 (0.47, 1.60)
0.650
0.86 (0.69, 1.06)
0.153
Model 3
0.76 (0.50, 1.16)
0.209
0.98 (0.84, 1.14)
0.779
1.02 (0.47, 2.19)
0.966
0.84 (0.65, 1.08)
0.184
Model 4
0.74 (0.49, 1.14)
0.171
0.96 (0.83, 1.12)
0.610
0.98 (0.45, 2.11)
0.956
0.82 (0.64, 1.07)
0.146
Model 5
0.65 (0.30, 1.40)
0.272
0.96 (0.76, 1.21)
0.718
1.01 (0.29, 3.53)
0.987
0.79 (0.56, 1.12)
0.194
PD
Model 1
0.59 (0.48, 0.73)
<0.001
1.24 (1.04, 1.47)
0.016
0.71 (0.52, 0.96)
0.027
1.16 (0.91, 1.47)
0.235
Model 2
0.71 (0.58, 0.87)
0.001
1.22 (1.02, 1.45)
0.024
0.82 (0.61, 1.11)
0.201
1.15 (0.90, 1.46)
0.260
Model 3
0.72 (0.58, 0.88)
0.002
1.14 (0.96, 1.36)
0.140
0.82 (0.61, 1.11)
0.209
1.07 (0.84, 1.37)
0.567
Model 4
0.74 (0.60, 0.91)
0.005
1.14 (0.96, 1.36)
0.147
0.85 (0.63, 1.15)
0.297
1.07 (0.84, 1.36)
0.602
Model 5
0.71 (0.55, 0.91)
0.006
1.08 (0.88, 1.31)
0.479
0.87 (0.60, 1.26)
0.575
1.03 (0.78, 1.37)
0.829
Model 1: Unadjusted.Model 2: Adjusted for age and sex.Model 3: Model 2 + DM + CVD history.Model 4: Model 3 + dialysis duration.Model 5: Model 4 + uric acid + bicarbonate + hemoglobin + cholesterol + triglyceride + LDL-c + phosphorus + calcium + iPTH.CVD: Cardiovascular disease; DM: Diabetes; HD: Hemodialysis; HR: Hazard ratio; iPTH: Intact parathyroid hormone; LDL-c: Low-density lipoprotein cholesterol; PD: Peritoneal dialysis.
Discussion
Traditionally, hyperkalemia is associated with adverse outcomes such as hospitalization events, CV death, and AC death.[5,7,14-16]
In our analyses, the lowest risk for mortality was at 4.0 mmol/L serum potassium. When potassium was <4.0 mmol/L, every 1.0 mmol/L decrease had a 26% increase in the risk of AC mortality. Consistently, multiple studies suggest that the optimal range is 4.0 to 5.0 mmol/L for potassium.[7,9] albumin , and other nutrition-related indicators, suggesting that potassium is a proxy of nutritional status.[17,18] albumin .[4,11] [16] [9] [5,7]
Interestingly, it is notable that there was an attenuation in the association of potassium and mortality when albumin was introduced into the extended adjusted Cox regression model in the overall population in this study, suggesting that albumin substantially affects their relationship. In line with this finding, a previous study showed that there was no significant association between serum potassium levels and mortality after multivariate adjustment, including for nutritional status.[5] albumin confounds the true relationship between potassium and mortality, it could also suggest that albumin is a mediator of the effect of potassium on mortality. Further analysis showed that albumin was a significant mediator in the link between potassium and AC and CV mortalities, verifying that the relationship between potassium and mortality was significantly mediated by nutritional status in this study. Additionally, most of the population with hyperkalemia in this study had mild or moderate hyperkalemia, with a relevant low proportion of severe hyperkalemia. Thus, the benefits of improved nutritional status may overcome the adverse risks associated with mild to moderate hyperkalemia. Besides, there are emerging data showing that serum potassium and albumin are potential oxidative stress and inflammation markers.[19-21] et al [19] albumin concentration in peritonitis patients and that both are independent risk factors for fugal peritonitis, suggesting potassium and albumin play an antioxidative role and are biomarkers of oxidative stress and inflammation.
Additionally, a disparity in the associations of potassium in patients in different dialysis modalities was noted in the present study. Hypokalemia was associated with elevated risks of death in PD, while hyperkalemia seemed to have a trend with an increased risk of mortality in HD patients. These inverse relationships may be partly explained by the different distributions of potassium in the PD and HD populations. Serum potassium was significantly lower in PD patients than in HD patients. Hypokalemia was more prominent in PD patients, while hyperkalemia was more common in HD patients in the present study. This may be due to the 24-hour continuous clearance of solutes, during which potassium shifts from extracellular to intracellular compartments driven by glucose absorption and subsequent insulin excretion during PD, and inadequate dialytic potassium removal during HD.[12] albumin were relatively lower in PD patients than in HD patients. Therefore, it is reasonable to believe that PD patients were in worse nutritional status than HD patients, as low dietary intake and malnutrition are the predominant causes of hypokalemia in dialysis patients, even surpassing medication use. Consistent with a previous study, energy intake and subsequent serum albumin and potassium levels were lower in PD patients than in HD patients, suggesting that dietary behavior and nutritional intake are worse in PD patients than in HD patients.[22] [23] [11,24]
Therefore, recognizing the prevalence and risk of hypokalemia for mortality and managing this subpopulation of dialysis patients are of great clinical importance. Recently, the treatment of hypokalemia has generally depended on the severity and presence of electrocardiograph abnormalities or symptoms, and thresholds for treatment initiation have not been defined. Our analyses showed that the lowest risk for mortality was at 4.0 mmol/L, and multiple studies suggest that the optimal range is 4.0 to 5.0 mmol/L[5,7]
The strengths of our study include a large prospective cohort of nationwide dialysis patients for whom serum potassium data were available. Additionally, we examined models with and without adjustment for serum albumin , which is a major marker of nutritional status and inflammation, to evaluate their underlying relationships. However, there are several limitations of our study. First, data on the dietary intake of potassium and protein were not collected in this study, and other nutritional and inflammation parameters were not available. Second, we only enrolled continuous ambulatory PD and HD patients; thus, our findings may not be representative of those treated with other modalities due to the differences in excretion mechanism and amount of excreted potassium. Third, the analysis was based only on baseline data rather than changes in these parameters. Fourth, although we tried to adjust for many relevant confounders in the analysis, we could not eliminate all possibilities of confounding.
In conclusion, serum potassium had a U-shaped association with AC mortality. The nadir for AC mortality risk was estimated to be at a potassium concentration of 4.0 mmol/L. A decrease in serum potassium below 4.0 mmol/L was associated with elevated risks of AC mortality in overall and PD patients, while the association disappeared above this level. However, serum potassium was not found to be associated with AC mortality in HD patients. Furthermore, serum albumin was demonstrated to be a mediator in the association of potassium and mortality in the overall population, suggesting that the relationship between potassium and mortality was mediated by albumin .
Acknowledgments
We thank all our colleagues at Sun Yat-Sen University for their excellent data collection and analysis.
Funding
This work was supported by grants from the National Natural Science Foundation of China (Nos. 82200820, 81970599 and 82170737), Guangzhou Science and Technology Project (No. 202201011483), Key Laboratory of National Health Commission, and Key Laboratory of Nephrology, Guangdong Province, Guangzhou, China (Nos. 2002B60118 and 2020B1212060028), and 5010 Clinical Program of Sun Yat-Sen University (No. 2017007), and National Key Research and Development Project of China (No. 2021YFC2501302).
Conflicts of interest
None.
Data availability statement
No additional data available.
References
1. Kashihara N, Kohsaka S, Kanda E, Okami S, Yajima T. Hyperkalemia in real-world patients under continuous medical care in Japan. Kidney Int Rep 2019;4:1248–1260. doi: 10.1016/j.ekir.2019.05.018.
2. Bem D, Sugrue D, Wilding B, Zile I, Butler K, Booth D, et al. The effect of hyperkalemia and long inter-dialytic interval on morbidity and mortality in patients receiving hemodialysis: a systematic review. Ren Fail 2021;43:241–254. doi: 10.1080/0886022X.2020.1871012.
3. Sarafidis PA, Blacklock R, Wood E, Rumjon A, Simmonds S, Fletcher-Rogers J, et al. Prevalence and factors associated with hyperkalemia in predialysis patients followed in a low-clearance clinic. Clin J Am Soc Nephrol 2012;7:1234–1241. doi: 10.2215/CJN.01150112.
4. Huang NY, Liu YY, Yu JW, Xu YW, Zheng XH, Zhang DH, et al. Current status of hyperkalemia in dialysis patients in China (in Chinese). Natl Med J China 2021;101:3466–3471. doi: 10.3760/cma.j.cn112137-20210802-01710.
5. Tanaka K, Saito H, Iwasaki T, Oda A, Watanabe S, Kanno M, et al. Association between serum potassium levels and adverse outcomes in chronic kidney disease: the Fukushima CKD cohort study. Clin Exp Nephrol 2021;25:410–417. doi: 10.1007/s10157-020-02010-7.
6. Elliott AB, Soliman KMM, Ullian ME. Hyperkalemia in chronic peritoneal dialysis patients. Ren Fail 2022;44:217–223. doi: 10.1080/0886022X.2022.2032151.
7. Brookes EM, Snider J, Hart GK, Robbins R, Power DA. Serum potassium abnormalities in chronic kidney disease: prevalence, patient characteristics and clinical outcomes. Intern Med J 2021;51:1906–1918. doi: 10.1111/imj.14970.
8. Hayes J, Kalantar-Zadeh K, Lu JL, Turban S, Anderson JE, Kovesdy CP. Association of hypo- and hyperkalemia with disease progression and mortality in males with chronic kidney disease: the role of race. Nephron Clin Pract 2012;120:c8–c16. doi: 10.1159/000329511.
9. Korgaonkar S, Tilea A, Gillespie BW, Kiser M, Eisele G, Finkelstein F, et al. Serum potassium and outcomes in CKD: Insights from the RRI-CKD cohort study. Clin J Am Soc Nephrol 2010;5:762–769. doi: 10.2215/CJN.05850809.
10. Clase CM, Carrero JJ, Ellison DH, Grams ME, Hemmelgarn BR, Jardine MJ, et al. Potassium homeostasis and management of dyskalemia in kidney diseases: conclusions from a kidney disease: improving global outcomes (KDIGO) controversies conference. Kidney Int 2020;97:42–61. doi: 10.1016/j.kint.2019.09.018.
11. Xu Q, Xu F, Fan L, Xiong L, Li H, Cao S, et al. Serum potassium levels and its variability in incident peritoneal dialysis patients: associations with mortality. PLoS One 2014;9:e86750. doi: 10.1371/journal.pone.0086750.
12. Xia X, He F, Wu X, Peng F, Huang F, Yu X. Relationship between serum uric acid and all-cause and cardiovascular mortality in patients treated with peritoneal dialysis. Am J Kidney Dis 2014;64:257–264. doi: 10.1053/j.ajkd.2013.08.027.
13. Cheung MW. metaSEM: An R package for meta-analysis using structural equation modeling. Front Psychol 2015;5:1521. doi: 10.3389/fpsyg.2014.01521.
14. Palaka E, Grandy S, Darlington O, McEwan P, van Doornewaard A. Associations between serum potassium and adverse clinical outcomes: a systematic literature review. Int J Clin Pract 2020;74:e13421. doi: 10.1111/ijcp.13421.
15. Kovesdy CP, Matsushita K, Sang Y, Brunskill NJ, Carrero JJ, Chodick G, et al. Serum potassium and adverse outcomes across the range of kidney function: a CKD prognosis consortium meta-analysis. Eur Heart J 2018;39:1535–1542. doi: 10.1093/eurheartj/ehy100.
16. Karaboyas A, Zee J, Brunelli SM, Usvyat LA, Weiner DE, Maddux FW, et al. Dialysate potassium, serum potassium, mortality, and arrhythmia events in hemodialysis: results from the dialysis outcomes and practice patterns study (DOPPS). Am J Kidney Dis 2017;69:266–277. doi: 10.1053/j.ajkd.2016.09.015.
17. Vega LB, Galabia ER, Silva JBD, Gonzalez MB, Fresnedo GF, Haces CP, et al. Epidemiology of hyperkalemia in chronic kidney disease. Nefrologia (Engl Ed) 2019;39:277–286. doi: 10.1016/j.nefro.2018.11.011.
18. Szeto CC, Chow KM, Kwan BC, Leung CB, Chung KY, Law MC, et al. Hypokalemia in Chinese peritoneal dialysis patients: prevalence and prognostic implication. Am J Kidney Dis 2005;46:128–135. doi: 10.1053/j.ajkd.2005.03.015.
19. Liu L, Xie K, Yin M, Chen X, Chen B, Ke J, et al. Serum potassium,
albumin and vitamin B12 as potential oxidative stress markers of fungal peritonitis. Ann Med 2021;53:2132–2141. doi: 10.1080/07853890.2021.1999489.
20. Eckart A, Struja T, Kutz A, Baumgartner A, Baumgartner T, Zurfluh S, et al. Relationship of nutritional status, inflammation, and serum
albumin levels during acute illness: a prospective study. Am J Med 2020;133:713–722.e2. doi: 10.1016/j.amjmed.2019.10.031.
21. Di A, Xiong S, Ye Z, Malireddi RKS, Kometani S, Zhong M, et al. The TWIK2 potassium efflux channel in macrophages mediates NLRP3 inflammasome-induced inflammation. Immunity 2018;49:56–65.e4. doi: 10.1016/j.immuni.2018.04.032.
22. Kim SM, Kang BC, Kim HJ, Kyung MS, Oh HJ, Kim JH, et al. Comparison of hemodialysis and peritoneal dialysis patients’ dietary behaviors. BMC Nephrol 2020;21:91. doi: 10.1186/s12882-020-01744-6.
23. Banshodani M, Kawanishi H, Moriishi M, Shintaku S, Tsuchiya S. Association between dialysis modality and infectious diseases: peritoneal dialysis versus hemodialysis. Blood Purif 2021;50:370–379. doi: 10.1159/000511041.
24. Bowling CB, Pitt B, Ahmed MI, Aban IB, Sanders PW, Mujib M, et al. Hypokalemia and outcomes in patients with chronic heart failure and chronic kidney disease: findings from propensity-matched studies. Circ Heart Fail 2010;3:253–260. doi: 10.1161/CIRCHEARTFAILURE.109.899526.
