http://informahealthcare.com/rnf ISSN: 0886-022X (print), 1525-6049 (electronic) Ren Fail, 2014; 36(5): 748–754 ! 2014 Informa Healthcare USA, Inc. DOI: 10.3109/0886022X.2014.884379
CLINICAL STUDY
Potassium metabolism in continuous ambulatory peritoneal dialysis patients Hong-Lei Yu, Xin-Hong Lu, Chun-Yan Su, Wen Tang, and Tao Wang
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Division of Nephrology, Peking University Third Hospital, Beijing, P.R. China
Abstract
Keywords
Background: Hypokalemia is common and may have contributed to the poor clinical outcome in peritoneal dialysis (PD) patients. In this study, we made a detailed investigation on the potassium metabolism in continuous ambulatory peritoneal dialysis (CAPD) patients and tried to find out the possible factors associated with the high prevalence of hypokalemia in PD patients. Methods: A cross-sectional survey in 243 clinically stable CAPD patients was made in our PD center in 2010. Patients were divided into four groups according to whether they were anuric or not and different dialysis regimens. Patients’ demographic data and data on potassium metabolism including dietary potassium intakes, residual renal potassium, and peritoneal dialysis potassium removal were collected. Results: The average potassium intake in our 243 PD patients was 32.1 ± 11.1 mmol/day. The total potassium removal was significantly higher in non-anuric patients as compared to anuric patients (33.2 ± 9.1 vs. 23.0 ± 4.7 mmol/day for 3 exchanges per day and 35.2 ± 8.9 vs. 28.6 ± 6.3 mmol/day for 4 exchanges per day, respectively, p50.01) and in anuric patients dialyzed with 4 exchanges per day as compared to anuric patients dialyzed with 3 exchanges per day (28.6 ± 6.3 vs. 23.0 ± 4.7 mmol/L, p50.05). Compared to non-anuric patients dialyzed with 3 exchanges per day, serum potassium level was significantly lower (4.1 ± 0.7 vs. 4.5 ± 0.7 mmol/L, p50.05) while the prevalence of hypokalemia was significantly higher (22.2% vs. 9.3%, p50.05) in non-anuric patients that dialyzed with 4 exchanges per day. There was a strong correlation between renal potassium removal and renal urea Kt/V (R2 linear ¼ 0.645, p50.05). In a linear multiregression analysis, dietary potassium intake, intracellular water (ICW) significantly positively predicted serum potassium level while dialysis exchanges, residual renal function (RRF), D/P potassium all significantly negatively predicted serum potassium levels. Conclusions: Our study suggested that if potassium intake was limited in PD patients, we should be aware of the risk of hypokalemia with high doses of PD when patients have good RRF. Our study also suggested that potassium removal in PD patients may not necessarily reflect potassium intake even if serum potassium is normal, the effect of ICW should be considered when evaluating potassium homeostasis.
Continuous ambulatory peritoneal dialysis, hypokalemia, intracellular water, potassium metabolism, residual renal function
Introduction The prevalence of hypokalemia among peritoneal dialysis (PD) patients has been found to vary from 10% to 36%.1–3 The hazards of hypokalemia are multiple. Hypokalemia can cause cardiac problem, for example, arrhythmias with decreased cardiac output, tachycardia, or bradycardia (depending on the type of arrhythmia), weak peripheral pulses, and orthostatic hypotension.4 It can be associated with poor clinical outcomes, including malnutrition, severe comorbidity, and decreased patient survival.5 The combination of hypokalemia and malnutrition may affect gastrointestinal motility, resulting in bacterial overgrowth and causing peritonitis through transmural migration of enteric organisms.6
Address correspondence to Tao Wang, Division of Nephrology, Peking University Third Hospital, 49 North Garden Road, Beijing 1000191, P.R. China. Tel: +8610 82268890; E-mail: [email protected]
History Received 12 October 2013 Revised 9 December 2013 Accepted 28 December 2013 Published online 11 February 2014
It is well known that in normal subject, the autoregulation of the kidney plays important roles in preventing hypokalemia and hyperkalemia.7,8 In PD patients, even with the presence of residual renal function (RRF), the autoregulation capacity in potassium metabolism is significantly impaired due to significant loss of potassium through dialysis which is not autoregulated.8,9 However, the dialysis regimen in PD is usually fixed for a defined patient if the clinical condition is stable. Most PD program applied 3–4 exchanges per day for the majority of PD patients. How this fixed dialysis protocol affects the potassium metabolism in PD patients is not well studied. Furthermore, although PD preserves RRF better as compared to hemodialysis, the impact of RRF on the potassium metabolism in PD is not well studied either. Therefore, in this study, we made a detailed investigation on the potassium metabolism in our continuous ambulatory peritoneal dialysis (CAPD) patients and tried to find out the
Potassium metabolism in PD
DOI: 10.3109/0886022X.2014.884379
possible factors that may be associated with the high prevalence of hypokalemia in PD patients.
Methods
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Patient selection A cross-sectional survey on clinically stable CAPD patients who had been on PD for more than 3 months was made in our PD center in 2010. Patients’ demographic data and data on potassium metabolism including dietary potassium intakes, residual renal potassium removal and PD potassium removal were collected. Patients were divided into four groups according to whether they were anuric or not and different dialysis regimens: anuric patients dialyzed with 3 exchanges per day group (6 L/day), non-anuric patients dialyzed with 3 exchanges per day group (6 L/day), anuric patients dialyzed with 4 exchanges per day group (8 L/day) and non-anuric patients dialyzed with 4 exchanges per day group (8 L/day). The exclusion criteria were: (1) unable to provide detailed dietary records and/or dialysis adequacy data; (2) unwilling to participate in this study; (3) peritonitis in the previous 3 months; (4) diarrhea or vomiting in the previous 3 months; (5) dialysis protocol change in the previous month; (6) patients dialyzed with less than 3 exchanges per day or more than 4 exchanges per day; (7) patients who were using diuretics. The patients were asked to stop ACE Inhibitor or ARB for 24 h before the dialysis adequacy test if they were using these medications. Two hundred and forty-three PD patients who met the criteria were included in this study. The study protocol was approved by the ethic committee of Peking University and all patients provided their written informed consents. Demographic data and clinical characteristics Data collected included demographics (including age, gender, height, and weight), primary disease, serum biochemistries, duration of dialysis, and dialysis prescription. Body mass index (BMI) was calculated as weight (in kg)/height2 (in m2). All other measurements were carried out using routine procedures in the clinic. Dietary potassium intake Estimation of dietary potassium intake was undertaken by a dedicated dietitian at the outpatient clinic, using selfcompleted 3-day food diaries right before coming to the clinic. Dietary intakes, including dietary protein intake (DPI) and dietary energy intake (DEI), as well as dietary contents including potassium, were all calculated using computer software based on the National Food Content Table (Information Management System, Peking University, Beijing, China). Before dietary assessment, each patient received intensive education by the dietitian on how to record dietary intake with the help of portion-sized food models.10 Potassium removal measurement Dialysis adequacy assessment was performed in all patients with 24 h dialysate and urine samples right before the clinic and the blood sample was taken at the clinic. In this study,
749
both the renal and peritoneal Kt/V urea11 as well as the 24 h renal and peritoneal potassium removal was calculated. Peritoneal potassium removal (mmol) ¼ 24 h dialysate potassium concentration (mmol/L) volume of 24 h dialysis effluent (L). Renal potassium removal (mmol) ¼ 24 h urine potassium concentration (mmol/L) 24 h urine volume (L). Total potassium removal ¼ peritoneal potassium removal (mmol) + renal potassium removal (mmol) Measurement of volume status As volume status would affect the potassium pool in the body, the patients’ volume status were assessed by multiplefrequency bioelectrical impedance spectroscopy (MF-BIS) analysis (BCM, Fresenius Medical Care, Bad Homburg, Germany) in this study. Briefly, MF-BIS analysis was performed with the patient in a supine position to ensure equilibration of fluid. Electrodes were placed on the wrist and ankle. Patients were measured without dialysate in the peritoneum and the result was regarded as valid only if the operative quality score was over 80%. The values of overhydration (OH), extracellular water (ECW), intracellular water (ICW), total body water (TBW), and ECW/ICW ratio (E/I) were recorded. OH means the difference between the normal expected ECW and the measured ECW. Normal ECW can be determined for a given weight and body composition. Our previous study suggested that OH below 2.0 is a reasonable cut-off value as normal hydration in PD patients.12 Measurements of SGA Subjective global assessment (SGA) was performed by a dedicated dietician in our program. The method has been described previously.13 Patients were classified into well nourished (SGA ¼ A) or malnourished (SGA ¼ B or C). Statistical analysis Statistical analysis was performed using SPSS for Windows software, version 16.0 (SPSS Inc., Chicago, IL). Results are expressed as mean ± SD unless otherwise specified. Comparisons between groups were performed using the 2 test, Student t-test, one-way ANOVA or Pearsons’ correlation analysis, as appropriate. A linear multiregression analysis was applied to study the possible factors that affected the serum potassium level. A two-tailed p value less than 0.05 was considered as statistically significant.
Results A total of 243 PD patients were included in this study. The patients’ basic characteristics were depicted in Table 1. There were no significant differences in age and the prevalence of diabetics among the groups. The prevalence of malnutrition in anuric patients was significantly higher as compared to their non-anuric counterparts both in 3 exchanges per day and 4 exchanges per day (28.9% vs. 19.8% for 3 exchanges per day and 29.8% vs. 17.8% for 4 exchanges per day, respectively, p50.01 for both). The prevalence of hypokalemia in the non-anuric 4 exchanges per day was
750
H.-L. Yu et al.
Ren Fail, 2014; 36(5): 748–754
Table 1. Patients’ demographics and clinic indices in the 243 CAPD patients. Three exchanges per day (6 L/day) Total (n ¼ 243)
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Gender (M/F) DM (%) Malnutrition (%) Age (years) Weight (kg) Height (cm) BMI (kg/m2) TPD (months) HypoK (%)
123/120 29.6 23.9 60.8 ± 14.9 61.9 ± 11.2 162.4 ± 11.0 23.7 ± 3.6 37.4 ± 26.6 14.8
Anuric (n ¼ 45) c
12/33 26.7 28.9 60.5 ± 15.8 55.5 ± 9.3c 158.9 ± 9.3c 22.0 ± 3.4 47.4 ± 29.8 17.8
Four exchanges per day (8 L/day)
Non-anuric (n ¼ 86)
Anuric (n ¼ 67)
Non-anuric (n ¼ 45)
48/38 27.9 19.8a 58.3 ± 15.9 61.4 ± 10.5 162.4 ± 7.8 23.2 ± 3.4 29.1 ± 21.3a 9.3
37/30 40.3 29.8 62.8 ± 13.9 63.5 ± 11.4 163.9 ± 15.8 23.9 ± 4.4b 45.0 ± 28.0 14.9
26/19 42.2 17.8a 62.9 ± 13.1 66.8 ± 11.3 163.3 ± 8.6 25.0 ± 3.8b 32.1 ± 24.6a 22.2b
Notes: DM, diabetes mellitus; TPD, time on PD; HypoK, prevalence of hypokalemia; CAPD, continuous ambulatory peritoneal dialysis. a Compared to anuric patients with the same number of exchange per day, p50.01. b Compared to non-anuric patients with 3 exchanges per day, p50.05. c Compared to all the other three groups, p50.05.
Figure 1. Distribution of daily dietary potassium intake in the 243 CAPD patients. The average potassium intake in our 243 PD patients was 32.1 ± 11.1 mmol/day.
significantly higher than that of the non-anuric 3 exchanges per day (22.2% vs. 9.3%, p50.05). The distribution of potassium intake in these patients is shown in Figure 1. The average potassium intake in our 243 PD patients was 32.1 ± 11.1 mmol/day. The data of potassium intake and removal in the 243 CAPD patients are shown in Table 2. Serum potassium level was significantly lower in non-anuric patients that dialyzed with 4 exchanges per day as compared to non-anuric patients dialyzed with 3 exchanges per day (4.1 ± 0.7 vs. 4.5 ± 0.7 mmol/L, p50.05), whereas the difference between the other groups was not statistically different. Peritoneal potassium removal was significantly higher in patients with 4 exchanges as compared to 3 exchanges no matter patients were anuric or non-anuric (28.6 ± 6.3 vs. 23.0 ± 4.7 mmol/day for anuric and 28.0 ± 4.9 vs. 23.3 ± 4.2 mmol/day for non-anuric, respectively, p50.05 for both). Therefore, the total potassium removal was
significantly higher in non-anuric patients as compared to anuric patients (33.2 ± 9.1 vs. 23.0 ± 4.7 mmol/day for 3 exchanges per day and 35.2 ± 8.9 vs. 28.6 ± 6.3 mmol/day for 4 exchanges per day, respectively, p50.01 for both). Besides, anuric patients dialyzed with 4 exchanges per day had higher total potassium removal as compared to anuric patients dialyzed with 3 exchanges per day (28.6 ± 6.3 vs. 23.0 ± 4.7 mmol/day, p50.05). The 24 h D/P potassium was significantly lower in anuric 4 exchanges per day group as compared to 3 exchanges per day group (0.78 ± 0.10 vs. 0.82 ± 0.10, p50.05). Serum potassium concentration was not significantly correlated with urine potassium concentration (R ¼ 0.091, p40.05), whereas serum potassium concentration was significantly correlated with the dialysate potassium concentration (R ¼ 0.766, p50.001). Patients’ volume status is also presented in Table 2. There was no significant difference in OH value among the four
Potassium metabolism in PD
DOI: 10.3109/0886022X.2014.884379
751
Table 2. Potassium homeostasis and volume status in the 243 CAPD patients. Three exchanges per day (6 L/day)
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Serum potassium (mmol/L) Potassium intake (mmol/day) D/P potassium DPR (mmol/day) RPR (mmol/day) TPR (mmol/day) ECW (L) ICW (L) TBW (L) OH (L)
Four exchanges per day (8 L/day)
Total (n ¼ 243)
Anuric
Non-anuric
Anuric
Non-anuric
4.3 ± 0.7 32.1 ± 11.1 0.81 ± 0.1 25.6 ± 5.7 4.8 ± 7.0 30.4 ± 8.6 15.6 ± 3.4 15.0 ± 4.3 30.6 ± 7.0 2.4 ± 1.9
4.2 ± 0.8 30.2 ± 10.0 0.82 ± 0.10 23.0 ± 4.7 0 23.0 ± 4.7 13.8 ± 3.1 13.0 ± 4.3 27.0 ± 6.9 1.9 ± 1.2
4.5 ± 0.7 34.9 ± 11.1a 0.83 ± 0.10 23.3 ± 4.2 10.0 ± 7.4a 33.2 ± 9.1a 15.6 ± 3.3a 16.1 ± 4.1a 31.8 ± 6.8a 2.3 ± 1.9
4.3 ± 0.8 30.0 ± 12.2 0.78 ± 0.10b 28.6 ± 6.3b 0 28.6 ± 6.3b 16.1 ± 3.5b 14.4 ± 4.1 30.3 ± 7.2b 2.5 ± 2.2
4.1 ± 0.7b 31.5 ± 9.3 0.80 ± 0.09 28.0 ± 4.9b 7.2 ± 6.9a,b 35.2 ± 8.9a 16.4 ± 3.2 15.7 ± 4.2 32.2 ± 6.4 2.6 ± 1.9
Notes: D/P, 24 h dialysate versus serum potassium concentration ratio; DPR, 24 h dialysate potassium removal; RPR, 24 h renal potassium removal; TPR, 24 h total potassium removal; ECW, extracellular water; ICW, intracellular water; TBW, total body water; OH, overhydration. a Compared to anuric patients with the same number of exchange per day, p50.01. b Compared to the corresponding patients with three exchanges per day, p50.05.
Table 3. Dietary information, dialysis adequacy, and biochemistry data. Three exchanges per day (6 L/day)
DPI (g/kg/day) DEI (kcal/kg/day) PKt/V RKt/V TKt/V UF (mL) SNa (mmol/L) Salb (g/L) Sglu (mmol/L) SHCO3 (mmol/L)
Four exchanges per day (8 L/day)
Total (n ¼ 243)
Anuric
Non-anuric
Anuric
Non-anuric
0.80 ± 0.20 25.4 ± 6.5 1.71 ± 0.52 0.04 ± 0.06 1.75 ± 0.51 530.3 ± 387.7 138.8 ± 3.6 38.5 ± 4.7 6.3 ± 2.2 28.3 ± 3.3
0.78 ± 0.25 25.7 ± 6.6 1.82 ± 0.48 0 1.82 ± 0.48 656.9 ± 299.2 137.8 ± 4.8 38.4 ± 5.3 5.9 ± 1.4 28.8 ± 3.2
0.86 ± 0.23 26.8 ± 7.3 1.40 ± 0.37a 0.09 ± 0.05a 1.48 ± 0.37c 321.3 ± 342.0a 139.4 ± 3.2 39.3 ± 4.8 5.9 ± 2.2 28.3 ± 3.2
0.76 ± 0.22 23.7 ± 5.8 1.96 ± 0.56b 0 1.96 ± 0.56b 694.4 ± 378.0 138.3 ± 3.6 37.9 ± 4.1 6.6 ± 3.2 28.1 ± 3.5
0.79 ± 0.21 24.7 ± 4.9 1.82 ± 0.44b 0.06 ± 0.05a 1.89 ± 0.46 391.0 ± 58.9a 139.2 ± 2.9 38.2 ± 4.3 7.0 ± 2.7b 28.3 ± 3.1
Notes: DPI, dietary protein intake; DEI, dietary energy intake; PKt/V, peritoneal urea Kt/V, RKt/V, renal urea Kt/V; TKt/V, total urea Kt/V; UL, daily peritoneal ultrafiltration; SNa, serum sodium concentration; Salb, serum albumin concentration; Sglu, serum glucose concentration; SHCO3, serum bicarbonate. a Comparison of anuric and non-anuric patients with same exchange per day, p50.05. b Comparison of different exchange anuric or non-anuric patients, p50.05. c Compared to all the other three groups, p50.05.
groups. However, the ECW, ICW, and TBW were all significantly higher in non-anuric patients dialyzed with 3 exchanges per day as compared to anuric patients dialyzed with 3 exchanges per day. Anuric patients dialyzed with 4 exchanges per day also had significantly higher ECW and TBW as compared to anuric patients dialyzed with 3 exchanges per day (16.1 ± 3.5 vs. 13.8 ± 3.1 L and 30.3 ± 7.2 vs. 27.0 ± 6.9 L, respectively, p50.05 for both). There was no significant difference in volume status between anuric and non-anuric patients dialyzed with 4 exchanges per day. Table 3 showed patients’ dietary information, dialysis adequacy data, and biochemistry data. There were no significant differences in DPI, DEI, serum sodium concentration, serum albumin concentration, serum bicarbonate concentration among the four groups. However, plasma glucose level was significantly higher in non-anuric patients dialyzed with 4 exchanges per day as compared to patients dialyzed with 3 exchanges per day (7.0 ± 2.7 vs. 5.9 ± 2.2 mmol/L, p50.05). Peritoneal ultrafiltration was significantly lower in non-anuric patients as compared to their corresponding anuric patients. There was a strong
correlation between renal potassium removal and renal urea Kt/V as shown in Figure 2 (R2 linear ¼ 0.645, p50.05). In a linear multiregression analysis, dietary potassium intake, ICW significantly positively predict serum potassium level while dialysis exchanges, RRF, D/P potassium all significantly negatively predicted serum potassium levels (Table 4). Plasma glucose concentration, sodium concentration, albumin concentration, and ECW did not have significant impact on serum potassium concentration.
Discussion In this study, we found that the actual potassium intake in our CAPD patients was significantly lower than that was recommended in the literature and hypokalemia was common especially in those non-anuric patients dialyzed with 4 exchanges per day. Potassium is an important ion in our body and maintaining potassium homeostasis is thus critical. However, hypokalemia is not uncommon in PD patients and have adversely affected the outcome of PD patients. Previous reports showed that hypokalemia was an independent adverse prognostic indicator
752
H.-L. Yu et al.
Ren Fail, 2014; 36(5): 748–754
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Figure 2. A relationship between residual renal potassium removal and renal urea Kt/V. There was a strong correlation between renal potassium removal and renal urea Kt/V (R2 linear ¼ 0.645, p50.05).
Table 4. Factors that affect serum potassium concentration by linear regression analysis.
ICW D/P potassium Exchanges GFR Potassium intake
b
t
p
0.289 0.178 0.183 0.188 0.141
4.077 2.581 2.543 2.551 2.025
0.000 0.011 0.012 0.012 0.044
Notes: ICW, intracellular water; GFR, residual renal glomerular filtration rate calculated as the average renal urea and creatinine clearance.
in PD patients.5 It may lead to paralysis, leg cramps, and other debilitation and thus poor quality of life. It can also cause cardiac problems and peritonitis.4,6 In this study, the prevalence of hypokalemia in our CAPD patients was 14.8% and 22.2% of non-anuric patients who dialyzed with 4 exchanges per day developed hypokalemia, indicating that hypokalemia should not be neglected in the management of PD therapy. Although it is well known that in healthy subject serum potassium level is affected by potassium intake and removal (mainly regulated by kidney) as well as the distribution of potassium between extracellular space and intracellular space,14 detailed potassium homeostasis in PD has not been well investigated even with the fact that the renal function was significantly impaired and would be lost after 2–3 years on PD in most of the patients. Most studies attributed the hypokalemia in PD to inadequate potassium intake.1 However, it should be kept in mind that serum potassium level is not just simply decided by potassium intake but the results of the balance between intake and removal and the body distribution of potassium. The average potassium intake in our PD patients was only 32 mmol/L which is significantly lower than what has been
recommended in the literature15 but was very similar to some previous findings.16,17 Our data also showed that potassium intakes in our patients appeared to be a normal distribution as shown in Figure 1. This lower potassium intake pattern may be multifactorial including change of eating patterns (low potassium diet) or inadequate food intake in some PD patients due to gastrointestinal problems. A previous survey in elder Chinese population with normal kidney function found that the actual dietary potassium intake with traditional Chinese diet may be as low as 30–40 mmol/day, which is about 60% of that in most Americans but similar to what we found in this study. Furthermore, although a traditional Chinese diet is rich in vegetable and patients take normal or even high-potassium diet, the method of cookery involves extensive boiling and frying, resulting in a substantial reduction in potassium content in the dishes,5 this may lead to even overestimation of actual potassium intake in our study and should be kept in mind in prescribing dialysis therapy. It is well known that physiological potassium homeostasis is defined mainly by two different balances: the internal balance that represents potassium redistribution between the intracellular and the extracellular compartments, and the external balance that represents the potassium loss and uptake from the environment.18 Thus, lower than recommended potassium intake may not necessarily lead to hypokalemia. Our study clearly indicates that besides potassium intake, RRF, dialysis dose, D/P potassium and intracellular volume all significant affect serum potassium level. Kidney is an important organ in autoregulating serum potassium level.7,8 However, the lack of association between serum potassium concentration and urine potassium concentration suggested that renal autoregulation of potassium homeostasis in PD patients was impaired, despite the fact that high RRF was associated with higher potassium removal. On the other hand, the significant association between serum
Potassium metabolism in PD
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
DOI: 10.3109/0886022X.2014.884379
potassium concentration and dialysate potassium concentration indicate another non-natural autoregulation of potassium homeostasis dose exist in PD patients and this autoregulation conceivably plays important roles in maintaining potassium homeostasis in PD patients. Potassium is mainly removed by diffusion in PD,19 therefore altering plasma potassium concentration may significantly affect the dialysis removal of potassium. However, this extra autoregulation mechanism is also significantly affected by dialysis dose which is another strong factor affecting serum potassium level with higher dose higher potassium removal and is manually regulated. It is thus not surprising to see that hypokalemia could be easily developed in patient with RRF and inadequate potassium intake but receiving high dialysis dose. It is unfortunate that most PD program applied fixed dialysis regimen in PD, for example, 3–4 exchanges per day for the majority of PD patients, this practice may potentially predispose patients to hypokalemia although a supplementation with KCl-tablets may correct this disorder during PD therapy. Although the potassium losses through faeces, sweat and other body secretions were not measured in the present study and their contribution to serum potassium levels may be low, further studies are still needed to completely understand their roles in potassium balance in dialysis population. Our study indicated that, the dialysate prescription should be tailored to patients according to patients’ dietary intake, RRF as well peritoneal function status. It is interesting to note that in multiregression analysis, intracellular volume is the strongest predictor of serum potassium concentration in our patients. Although the accuracy of ECW and ICW measured by bioimpedance assessment is still an issue of debate, our results warrant further speculations. It is well known that although in PD the removed potassium is diffused from the extracellular space, potassium in the extracellular space is only a small part of the potassium pool in the body.20 Potassium is mainly present in the intracellular space where potassium concentration is about 30 times higher than that of potassium in the extracellular space.21 The distribution of potassium between the intracellular and extracellular fluid compartments is determined by many factors. The potassium shift through the cellular membrane follows passive diffusion from the intracellular to the extracellular compartment and, in the opposite direction, the active transport due to the Na–K ATPase transporter (Na–K pump).22 Factors that regulate Na–K pump in the active transport include catecholamine and insulin,23 while in the passive transport, potassium shift is regulated by the potassium concentration gradient between the intra- and the extracellular compartments. In fact, despite of removal via dialysis, insulin hormone, stimulated by the continuous glucose peritoneal infusion, could activate Na–K pump and generate an excess of potassium redistribution from extracellular into the intracellular compartment, leading to the decrease of plasma potassium.18 Moreover, acidosis could also reduce potassium intake by cells. There is a strong inverse relationship between blood pH and serum potassium concentration (SK), such that SK increases by about 0.8 mmol/L for each decline of 0.1 pH units.24 Subsequent equilibration
753
between the intra- and the extracellular compartments is driven by the potassium concentration gradient. Given the amount of potassium, the potassium concentration in intracellular compartment is determined by the distribution volume, ICW. It is thus not surprising to see that ICW significantly affects plasma potassium. Therefore, potassium removal in PD patients may not necessarily reflect potassium intake even if serum potassium is normal. In summary, our study suggested RRF plays an important role in potassium removal in CAPD patients. However, the regulation capacity in potassium metabolism of RRF was significantly impaired. Besides, the potassium level of CAPD patients also affected by intracellular volume, dietary intake, and dialysis dose. The appropriate PD dose should be given after detailed evaluation to avoid hypokalemia.
Acknowledgments The authors thank all the staff in the Division of Nephrology, Peking University Third Hospital.
Declaration of interest This study was partly supported by a grant from Beijing Municipal Science and Technology Foundation (D131100004713002) The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Khan AN, Bernardini J, Johnston JR, et al. Hypokalemia in peritoneal dialysis patients. Perit Dial Int. 1996;16(6):652. 2. Spital A, Sterns RH. Potassium supplementation via the dialysate in continuous ambulatory peritoneal dialysis. Am J Kidney Dis. 1985; 6(3):173–176. 3. Rostand SG. Profound hypokalemia in continuous ambulatory peritoneal dialysis. Arch Intern Med. 1983;143(2):377–378. 4. Burger CM. Emergency: hypokalemia. Am J Nurs. 2004; 104(11):61–65. 5. Szeto CC, Chow KM, Kwan BC, et al. Hypokalemia in Chinese peritoneal dialysis patients: prevalence and prognostic implication. Am J Kidney Dis. 2005;46(1):128–135. 6. Chuang YW, Shu KH, Yu TM, et al. Hypokalaemia: an independent risk factor of Enterobacteriaceae peritonitis in CAPD patients. Nephrol Dial Transplant. 2009;24(5):1603–1608. 7. Rippe B, Venturoli D. Optimum electrolyte composition of a dialysis solution. Perit Dial Int. 2008;28(3):131–136. 8. Alcazar AR. Electrolyte and acid-base balance disorders in advanced chronic kidney disease. Nefrologia. 2008; 28(3):87–93. 9. Redaelli B, Bonoldi G, Di Filippo G, et al. Behaviour of potassium removal in different dialytic schedules. Nephrol Dial Transplant. 1998;13(6):35–38. 10. Lu XH, Su CY, Sun LH, et al. Implementing continuous quality improvement process in potassium management in peritoneal dialysis patients. J Ren Nutr. 2009;19(6):469–474. 11. Nolph KD, Moore HL, Twardowski ZJ, et al. Cross-sectional assessment of weekly urea and creatinine clearances in patients on continuous ambulatory peritoneal dialysis. ASAIO J. 1992; 38(3):139–142. 12. Luo YJ, Lu XH, Woods F, et al. Volume control in peritoneal dialysis patients guided by bioimpedance spectroscopy assessment. Blood Purif. 2011;31(4):296–302.
754
H.-L. Yu et al.
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
13. Enia G, Sicuso C, Alati G, et al. Subjective global assessment of nutrition in dialysis patients. Nephrol Dial Transplant. 1993; 8(10):1094–1098. 14. Rachoin JS, Weisberg LS. How should dialysis fluid be individualized for the chronic hemodialysis patient? Semin Dial. 2008; 21(3):223–225. 15. Beto JA, Bansal VK. Medical nutrition therapy in chronic kidney failure: integrating clinical practice guidelines. J Am Diet Assoc. 2004;104(3):404–409. 16. Kwok TC, Chan TY, Woo J. Relationship of urinary sodium/ potassium excretion and calcium intake to blood pressure and prevalence of hypertension among older Chinese vegetarians. Eur J Clin Nutr. 2003;57(2):299–304. 17. Chow KM, Szeto CC, Wong TY, et al. Risk factors for thiazideinduced hyponatraemia. QJM 2003;96(12):911–917. 18. Tziviskou E, Musso C, Bellizzi V, et al. Prevalence and pathogenesis of hypokalemia in patients on chronic peritoneal dialysis: one
Ren Fail, 2014; 36(5): 748–754
19. 20. 21. 22. 23. 24.
center’s experience and review of the literature. Int Urol Nephrol. 2003;35(3):429–434. Musso CG. Potassium metabolism in patients with chronic kidney disease. Part II: patients on dialysis (stage 5). Int Urol Nephrol. 2004;36(3):469–472. Edelman IS, Leibman J. Anatomy of body water and electrolytes. Am J Med. 1959;27:256–277. Ahmed J, Weisberg LS. Hyperkalemia in dialysis patients. Semin Dial. 2001;14(5):348–356. Ciandrini A, Severi S, Cavalcanti S, et al. Model-based analysis of potassium removal during hemodialysis. Artif Organs. 2009; 33(10):835–843. Clausen T, Everts ME. Regulation of the Na, K-pump in skeletal muscle. Kidney Int. 1989;35(1):1–13. Magner PO, Robinson L, Halperin RM, et al. The plasma potassium concentration in metabolic acidosis: a re-evaluation. Am J Kidney Dis. 1988;11(3):220–224.
CLINICAL STUDY
Potassium metabolism in continuous ambulatory peritoneal dialysis patients Hong-Lei Yu, Xin-Hong Lu, Chun-Yan Su, Wen Tang, and Tao Wang
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Division of Nephrology, Peking University Third Hospital, Beijing, P.R. China
Abstract
Keywords
Background: Hypokalemia is common and may have contributed to the poor clinical outcome in peritoneal dialysis (PD) patients. In this study, we made a detailed investigation on the potassium metabolism in continuous ambulatory peritoneal dialysis (CAPD) patients and tried to find out the possible factors associated with the high prevalence of hypokalemia in PD patients. Methods: A cross-sectional survey in 243 clinically stable CAPD patients was made in our PD center in 2010. Patients were divided into four groups according to whether they were anuric or not and different dialysis regimens. Patients’ demographic data and data on potassium metabolism including dietary potassium intakes, residual renal potassium, and peritoneal dialysis potassium removal were collected. Results: The average potassium intake in our 243 PD patients was 32.1 ± 11.1 mmol/day. The total potassium removal was significantly higher in non-anuric patients as compared to anuric patients (33.2 ± 9.1 vs. 23.0 ± 4.7 mmol/day for 3 exchanges per day and 35.2 ± 8.9 vs. 28.6 ± 6.3 mmol/day for 4 exchanges per day, respectively, p50.01) and in anuric patients dialyzed with 4 exchanges per day as compared to anuric patients dialyzed with 3 exchanges per day (28.6 ± 6.3 vs. 23.0 ± 4.7 mmol/L, p50.05). Compared to non-anuric patients dialyzed with 3 exchanges per day, serum potassium level was significantly lower (4.1 ± 0.7 vs. 4.5 ± 0.7 mmol/L, p50.05) while the prevalence of hypokalemia was significantly higher (22.2% vs. 9.3%, p50.05) in non-anuric patients that dialyzed with 4 exchanges per day. There was a strong correlation between renal potassium removal and renal urea Kt/V (R2 linear ¼ 0.645, p50.05). In a linear multiregression analysis, dietary potassium intake, intracellular water (ICW) significantly positively predicted serum potassium level while dialysis exchanges, residual renal function (RRF), D/P potassium all significantly negatively predicted serum potassium levels. Conclusions: Our study suggested that if potassium intake was limited in PD patients, we should be aware of the risk of hypokalemia with high doses of PD when patients have good RRF. Our study also suggested that potassium removal in PD patients may not necessarily reflect potassium intake even if serum potassium is normal, the effect of ICW should be considered when evaluating potassium homeostasis.
Continuous ambulatory peritoneal dialysis, hypokalemia, intracellular water, potassium metabolism, residual renal function
Introduction The prevalence of hypokalemia among peritoneal dialysis (PD) patients has been found to vary from 10% to 36%.1–3 The hazards of hypokalemia are multiple. Hypokalemia can cause cardiac problem, for example, arrhythmias with decreased cardiac output, tachycardia, or bradycardia (depending on the type of arrhythmia), weak peripheral pulses, and orthostatic hypotension.4 It can be associated with poor clinical outcomes, including malnutrition, severe comorbidity, and decreased patient survival.5 The combination of hypokalemia and malnutrition may affect gastrointestinal motility, resulting in bacterial overgrowth and causing peritonitis through transmural migration of enteric organisms.6
Address correspondence to Tao Wang, Division of Nephrology, Peking University Third Hospital, 49 North Garden Road, Beijing 1000191, P.R. China. Tel: +8610 82268890; E-mail: [email protected]
History Received 12 October 2013 Revised 9 December 2013 Accepted 28 December 2013 Published online 11 February 2014
It is well known that in normal subject, the autoregulation of the kidney plays important roles in preventing hypokalemia and hyperkalemia.7,8 In PD patients, even with the presence of residual renal function (RRF), the autoregulation capacity in potassium metabolism is significantly impaired due to significant loss of potassium through dialysis which is not autoregulated.8,9 However, the dialysis regimen in PD is usually fixed for a defined patient if the clinical condition is stable. Most PD program applied 3–4 exchanges per day for the majority of PD patients. How this fixed dialysis protocol affects the potassium metabolism in PD patients is not well studied. Furthermore, although PD preserves RRF better as compared to hemodialysis, the impact of RRF on the potassium metabolism in PD is not well studied either. Therefore, in this study, we made a detailed investigation on the potassium metabolism in our continuous ambulatory peritoneal dialysis (CAPD) patients and tried to find out the
Potassium metabolism in PD
DOI: 10.3109/0886022X.2014.884379
possible factors that may be associated with the high prevalence of hypokalemia in PD patients.
Methods
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Patient selection A cross-sectional survey on clinically stable CAPD patients who had been on PD for more than 3 months was made in our PD center in 2010. Patients’ demographic data and data on potassium metabolism including dietary potassium intakes, residual renal potassium removal and PD potassium removal were collected. Patients were divided into four groups according to whether they were anuric or not and different dialysis regimens: anuric patients dialyzed with 3 exchanges per day group (6 L/day), non-anuric patients dialyzed with 3 exchanges per day group (6 L/day), anuric patients dialyzed with 4 exchanges per day group (8 L/day) and non-anuric patients dialyzed with 4 exchanges per day group (8 L/day). The exclusion criteria were: (1) unable to provide detailed dietary records and/or dialysis adequacy data; (2) unwilling to participate in this study; (3) peritonitis in the previous 3 months; (4) diarrhea or vomiting in the previous 3 months; (5) dialysis protocol change in the previous month; (6) patients dialyzed with less than 3 exchanges per day or more than 4 exchanges per day; (7) patients who were using diuretics. The patients were asked to stop ACE Inhibitor or ARB for 24 h before the dialysis adequacy test if they were using these medications. Two hundred and forty-three PD patients who met the criteria were included in this study. The study protocol was approved by the ethic committee of Peking University and all patients provided their written informed consents. Demographic data and clinical characteristics Data collected included demographics (including age, gender, height, and weight), primary disease, serum biochemistries, duration of dialysis, and dialysis prescription. Body mass index (BMI) was calculated as weight (in kg)/height2 (in m2). All other measurements were carried out using routine procedures in the clinic. Dietary potassium intake Estimation of dietary potassium intake was undertaken by a dedicated dietitian at the outpatient clinic, using selfcompleted 3-day food diaries right before coming to the clinic. Dietary intakes, including dietary protein intake (DPI) and dietary energy intake (DEI), as well as dietary contents including potassium, were all calculated using computer software based on the National Food Content Table (Information Management System, Peking University, Beijing, China). Before dietary assessment, each patient received intensive education by the dietitian on how to record dietary intake with the help of portion-sized food models.10 Potassium removal measurement Dialysis adequacy assessment was performed in all patients with 24 h dialysate and urine samples right before the clinic and the blood sample was taken at the clinic. In this study,
749
both the renal and peritoneal Kt/V urea11 as well as the 24 h renal and peritoneal potassium removal was calculated. Peritoneal potassium removal (mmol) ¼ 24 h dialysate potassium concentration (mmol/L) volume of 24 h dialysis effluent (L). Renal potassium removal (mmol) ¼ 24 h urine potassium concentration (mmol/L) 24 h urine volume (L). Total potassium removal ¼ peritoneal potassium removal (mmol) + renal potassium removal (mmol) Measurement of volume status As volume status would affect the potassium pool in the body, the patients’ volume status were assessed by multiplefrequency bioelectrical impedance spectroscopy (MF-BIS) analysis (BCM, Fresenius Medical Care, Bad Homburg, Germany) in this study. Briefly, MF-BIS analysis was performed with the patient in a supine position to ensure equilibration of fluid. Electrodes were placed on the wrist and ankle. Patients were measured without dialysate in the peritoneum and the result was regarded as valid only if the operative quality score was over 80%. The values of overhydration (OH), extracellular water (ECW), intracellular water (ICW), total body water (TBW), and ECW/ICW ratio (E/I) were recorded. OH means the difference between the normal expected ECW and the measured ECW. Normal ECW can be determined for a given weight and body composition. Our previous study suggested that OH below 2.0 is a reasonable cut-off value as normal hydration in PD patients.12 Measurements of SGA Subjective global assessment (SGA) was performed by a dedicated dietician in our program. The method has been described previously.13 Patients were classified into well nourished (SGA ¼ A) or malnourished (SGA ¼ B or C). Statistical analysis Statistical analysis was performed using SPSS for Windows software, version 16.0 (SPSS Inc., Chicago, IL). Results are expressed as mean ± SD unless otherwise specified. Comparisons between groups were performed using the 2 test, Student t-test, one-way ANOVA or Pearsons’ correlation analysis, as appropriate. A linear multiregression analysis was applied to study the possible factors that affected the serum potassium level. A two-tailed p value less than 0.05 was considered as statistically significant.
Results A total of 243 PD patients were included in this study. The patients’ basic characteristics were depicted in Table 1. There were no significant differences in age and the prevalence of diabetics among the groups. The prevalence of malnutrition in anuric patients was significantly higher as compared to their non-anuric counterparts both in 3 exchanges per day and 4 exchanges per day (28.9% vs. 19.8% for 3 exchanges per day and 29.8% vs. 17.8% for 4 exchanges per day, respectively, p50.01 for both). The prevalence of hypokalemia in the non-anuric 4 exchanges per day was
750
H.-L. Yu et al.
Ren Fail, 2014; 36(5): 748–754
Table 1. Patients’ demographics and clinic indices in the 243 CAPD patients. Three exchanges per day (6 L/day) Total (n ¼ 243)
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Gender (M/F) DM (%) Malnutrition (%) Age (years) Weight (kg) Height (cm) BMI (kg/m2) TPD (months) HypoK (%)
123/120 29.6 23.9 60.8 ± 14.9 61.9 ± 11.2 162.4 ± 11.0 23.7 ± 3.6 37.4 ± 26.6 14.8
Anuric (n ¼ 45) c
12/33 26.7 28.9 60.5 ± 15.8 55.5 ± 9.3c 158.9 ± 9.3c 22.0 ± 3.4 47.4 ± 29.8 17.8
Four exchanges per day (8 L/day)
Non-anuric (n ¼ 86)
Anuric (n ¼ 67)
Non-anuric (n ¼ 45)
48/38 27.9 19.8a 58.3 ± 15.9 61.4 ± 10.5 162.4 ± 7.8 23.2 ± 3.4 29.1 ± 21.3a 9.3
37/30 40.3 29.8 62.8 ± 13.9 63.5 ± 11.4 163.9 ± 15.8 23.9 ± 4.4b 45.0 ± 28.0 14.9
26/19 42.2 17.8a 62.9 ± 13.1 66.8 ± 11.3 163.3 ± 8.6 25.0 ± 3.8b 32.1 ± 24.6a 22.2b
Notes: DM, diabetes mellitus; TPD, time on PD; HypoK, prevalence of hypokalemia; CAPD, continuous ambulatory peritoneal dialysis. a Compared to anuric patients with the same number of exchange per day, p50.01. b Compared to non-anuric patients with 3 exchanges per day, p50.05. c Compared to all the other three groups, p50.05.
Figure 1. Distribution of daily dietary potassium intake in the 243 CAPD patients. The average potassium intake in our 243 PD patients was 32.1 ± 11.1 mmol/day.
significantly higher than that of the non-anuric 3 exchanges per day (22.2% vs. 9.3%, p50.05). The distribution of potassium intake in these patients is shown in Figure 1. The average potassium intake in our 243 PD patients was 32.1 ± 11.1 mmol/day. The data of potassium intake and removal in the 243 CAPD patients are shown in Table 2. Serum potassium level was significantly lower in non-anuric patients that dialyzed with 4 exchanges per day as compared to non-anuric patients dialyzed with 3 exchanges per day (4.1 ± 0.7 vs. 4.5 ± 0.7 mmol/L, p50.05), whereas the difference between the other groups was not statistically different. Peritoneal potassium removal was significantly higher in patients with 4 exchanges as compared to 3 exchanges no matter patients were anuric or non-anuric (28.6 ± 6.3 vs. 23.0 ± 4.7 mmol/day for anuric and 28.0 ± 4.9 vs. 23.3 ± 4.2 mmol/day for non-anuric, respectively, p50.05 for both). Therefore, the total potassium removal was
significantly higher in non-anuric patients as compared to anuric patients (33.2 ± 9.1 vs. 23.0 ± 4.7 mmol/day for 3 exchanges per day and 35.2 ± 8.9 vs. 28.6 ± 6.3 mmol/day for 4 exchanges per day, respectively, p50.01 for both). Besides, anuric patients dialyzed with 4 exchanges per day had higher total potassium removal as compared to anuric patients dialyzed with 3 exchanges per day (28.6 ± 6.3 vs. 23.0 ± 4.7 mmol/day, p50.05). The 24 h D/P potassium was significantly lower in anuric 4 exchanges per day group as compared to 3 exchanges per day group (0.78 ± 0.10 vs. 0.82 ± 0.10, p50.05). Serum potassium concentration was not significantly correlated with urine potassium concentration (R ¼ 0.091, p40.05), whereas serum potassium concentration was significantly correlated with the dialysate potassium concentration (R ¼ 0.766, p50.001). Patients’ volume status is also presented in Table 2. There was no significant difference in OH value among the four
Potassium metabolism in PD
DOI: 10.3109/0886022X.2014.884379
751
Table 2. Potassium homeostasis and volume status in the 243 CAPD patients. Three exchanges per day (6 L/day)
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Serum potassium (mmol/L) Potassium intake (mmol/day) D/P potassium DPR (mmol/day) RPR (mmol/day) TPR (mmol/day) ECW (L) ICW (L) TBW (L) OH (L)
Four exchanges per day (8 L/day)
Total (n ¼ 243)
Anuric
Non-anuric
Anuric
Non-anuric
4.3 ± 0.7 32.1 ± 11.1 0.81 ± 0.1 25.6 ± 5.7 4.8 ± 7.0 30.4 ± 8.6 15.6 ± 3.4 15.0 ± 4.3 30.6 ± 7.0 2.4 ± 1.9
4.2 ± 0.8 30.2 ± 10.0 0.82 ± 0.10 23.0 ± 4.7 0 23.0 ± 4.7 13.8 ± 3.1 13.0 ± 4.3 27.0 ± 6.9 1.9 ± 1.2
4.5 ± 0.7 34.9 ± 11.1a 0.83 ± 0.10 23.3 ± 4.2 10.0 ± 7.4a 33.2 ± 9.1a 15.6 ± 3.3a 16.1 ± 4.1a 31.8 ± 6.8a 2.3 ± 1.9
4.3 ± 0.8 30.0 ± 12.2 0.78 ± 0.10b 28.6 ± 6.3b 0 28.6 ± 6.3b 16.1 ± 3.5b 14.4 ± 4.1 30.3 ± 7.2b 2.5 ± 2.2
4.1 ± 0.7b 31.5 ± 9.3 0.80 ± 0.09 28.0 ± 4.9b 7.2 ± 6.9a,b 35.2 ± 8.9a 16.4 ± 3.2 15.7 ± 4.2 32.2 ± 6.4 2.6 ± 1.9
Notes: D/P, 24 h dialysate versus serum potassium concentration ratio; DPR, 24 h dialysate potassium removal; RPR, 24 h renal potassium removal; TPR, 24 h total potassium removal; ECW, extracellular water; ICW, intracellular water; TBW, total body water; OH, overhydration. a Compared to anuric patients with the same number of exchange per day, p50.01. b Compared to the corresponding patients with three exchanges per day, p50.05.
Table 3. Dietary information, dialysis adequacy, and biochemistry data. Three exchanges per day (6 L/day)
DPI (g/kg/day) DEI (kcal/kg/day) PKt/V RKt/V TKt/V UF (mL) SNa (mmol/L) Salb (g/L) Sglu (mmol/L) SHCO3 (mmol/L)
Four exchanges per day (8 L/day)
Total (n ¼ 243)
Anuric
Non-anuric
Anuric
Non-anuric
0.80 ± 0.20 25.4 ± 6.5 1.71 ± 0.52 0.04 ± 0.06 1.75 ± 0.51 530.3 ± 387.7 138.8 ± 3.6 38.5 ± 4.7 6.3 ± 2.2 28.3 ± 3.3
0.78 ± 0.25 25.7 ± 6.6 1.82 ± 0.48 0 1.82 ± 0.48 656.9 ± 299.2 137.8 ± 4.8 38.4 ± 5.3 5.9 ± 1.4 28.8 ± 3.2
0.86 ± 0.23 26.8 ± 7.3 1.40 ± 0.37a 0.09 ± 0.05a 1.48 ± 0.37c 321.3 ± 342.0a 139.4 ± 3.2 39.3 ± 4.8 5.9 ± 2.2 28.3 ± 3.2
0.76 ± 0.22 23.7 ± 5.8 1.96 ± 0.56b 0 1.96 ± 0.56b 694.4 ± 378.0 138.3 ± 3.6 37.9 ± 4.1 6.6 ± 3.2 28.1 ± 3.5
0.79 ± 0.21 24.7 ± 4.9 1.82 ± 0.44b 0.06 ± 0.05a 1.89 ± 0.46 391.0 ± 58.9a 139.2 ± 2.9 38.2 ± 4.3 7.0 ± 2.7b 28.3 ± 3.1
Notes: DPI, dietary protein intake; DEI, dietary energy intake; PKt/V, peritoneal urea Kt/V, RKt/V, renal urea Kt/V; TKt/V, total urea Kt/V; UL, daily peritoneal ultrafiltration; SNa, serum sodium concentration; Salb, serum albumin concentration; Sglu, serum glucose concentration; SHCO3, serum bicarbonate. a Comparison of anuric and non-anuric patients with same exchange per day, p50.05. b Comparison of different exchange anuric or non-anuric patients, p50.05. c Compared to all the other three groups, p50.05.
groups. However, the ECW, ICW, and TBW were all significantly higher in non-anuric patients dialyzed with 3 exchanges per day as compared to anuric patients dialyzed with 3 exchanges per day. Anuric patients dialyzed with 4 exchanges per day also had significantly higher ECW and TBW as compared to anuric patients dialyzed with 3 exchanges per day (16.1 ± 3.5 vs. 13.8 ± 3.1 L and 30.3 ± 7.2 vs. 27.0 ± 6.9 L, respectively, p50.05 for both). There was no significant difference in volume status between anuric and non-anuric patients dialyzed with 4 exchanges per day. Table 3 showed patients’ dietary information, dialysis adequacy data, and biochemistry data. There were no significant differences in DPI, DEI, serum sodium concentration, serum albumin concentration, serum bicarbonate concentration among the four groups. However, plasma glucose level was significantly higher in non-anuric patients dialyzed with 4 exchanges per day as compared to patients dialyzed with 3 exchanges per day (7.0 ± 2.7 vs. 5.9 ± 2.2 mmol/L, p50.05). Peritoneal ultrafiltration was significantly lower in non-anuric patients as compared to their corresponding anuric patients. There was a strong
correlation between renal potassium removal and renal urea Kt/V as shown in Figure 2 (R2 linear ¼ 0.645, p50.05). In a linear multiregression analysis, dietary potassium intake, ICW significantly positively predict serum potassium level while dialysis exchanges, RRF, D/P potassium all significantly negatively predicted serum potassium levels (Table 4). Plasma glucose concentration, sodium concentration, albumin concentration, and ECW did not have significant impact on serum potassium concentration.
Discussion In this study, we found that the actual potassium intake in our CAPD patients was significantly lower than that was recommended in the literature and hypokalemia was common especially in those non-anuric patients dialyzed with 4 exchanges per day. Potassium is an important ion in our body and maintaining potassium homeostasis is thus critical. However, hypokalemia is not uncommon in PD patients and have adversely affected the outcome of PD patients. Previous reports showed that hypokalemia was an independent adverse prognostic indicator
752
H.-L. Yu et al.
Ren Fail, 2014; 36(5): 748–754
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
Figure 2. A relationship between residual renal potassium removal and renal urea Kt/V. There was a strong correlation between renal potassium removal and renal urea Kt/V (R2 linear ¼ 0.645, p50.05).
Table 4. Factors that affect serum potassium concentration by linear regression analysis.
ICW D/P potassium Exchanges GFR Potassium intake
b
t
p
0.289 0.178 0.183 0.188 0.141
4.077 2.581 2.543 2.551 2.025
0.000 0.011 0.012 0.012 0.044
Notes: ICW, intracellular water; GFR, residual renal glomerular filtration rate calculated as the average renal urea and creatinine clearance.
in PD patients.5 It may lead to paralysis, leg cramps, and other debilitation and thus poor quality of life. It can also cause cardiac problems and peritonitis.4,6 In this study, the prevalence of hypokalemia in our CAPD patients was 14.8% and 22.2% of non-anuric patients who dialyzed with 4 exchanges per day developed hypokalemia, indicating that hypokalemia should not be neglected in the management of PD therapy. Although it is well known that in healthy subject serum potassium level is affected by potassium intake and removal (mainly regulated by kidney) as well as the distribution of potassium between extracellular space and intracellular space,14 detailed potassium homeostasis in PD has not been well investigated even with the fact that the renal function was significantly impaired and would be lost after 2–3 years on PD in most of the patients. Most studies attributed the hypokalemia in PD to inadequate potassium intake.1 However, it should be kept in mind that serum potassium level is not just simply decided by potassium intake but the results of the balance between intake and removal and the body distribution of potassium. The average potassium intake in our PD patients was only 32 mmol/L which is significantly lower than what has been
recommended in the literature15 but was very similar to some previous findings.16,17 Our data also showed that potassium intakes in our patients appeared to be a normal distribution as shown in Figure 1. This lower potassium intake pattern may be multifactorial including change of eating patterns (low potassium diet) or inadequate food intake in some PD patients due to gastrointestinal problems. A previous survey in elder Chinese population with normal kidney function found that the actual dietary potassium intake with traditional Chinese diet may be as low as 30–40 mmol/day, which is about 60% of that in most Americans but similar to what we found in this study. Furthermore, although a traditional Chinese diet is rich in vegetable and patients take normal or even high-potassium diet, the method of cookery involves extensive boiling and frying, resulting in a substantial reduction in potassium content in the dishes,5 this may lead to even overestimation of actual potassium intake in our study and should be kept in mind in prescribing dialysis therapy. It is well known that physiological potassium homeostasis is defined mainly by two different balances: the internal balance that represents potassium redistribution between the intracellular and the extracellular compartments, and the external balance that represents the potassium loss and uptake from the environment.18 Thus, lower than recommended potassium intake may not necessarily lead to hypokalemia. Our study clearly indicates that besides potassium intake, RRF, dialysis dose, D/P potassium and intracellular volume all significant affect serum potassium level. Kidney is an important organ in autoregulating serum potassium level.7,8 However, the lack of association between serum potassium concentration and urine potassium concentration suggested that renal autoregulation of potassium homeostasis in PD patients was impaired, despite the fact that high RRF was associated with higher potassium removal. On the other hand, the significant association between serum
Potassium metabolism in PD
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
DOI: 10.3109/0886022X.2014.884379
potassium concentration and dialysate potassium concentration indicate another non-natural autoregulation of potassium homeostasis dose exist in PD patients and this autoregulation conceivably plays important roles in maintaining potassium homeostasis in PD patients. Potassium is mainly removed by diffusion in PD,19 therefore altering plasma potassium concentration may significantly affect the dialysis removal of potassium. However, this extra autoregulation mechanism is also significantly affected by dialysis dose which is another strong factor affecting serum potassium level with higher dose higher potassium removal and is manually regulated. It is thus not surprising to see that hypokalemia could be easily developed in patient with RRF and inadequate potassium intake but receiving high dialysis dose. It is unfortunate that most PD program applied fixed dialysis regimen in PD, for example, 3–4 exchanges per day for the majority of PD patients, this practice may potentially predispose patients to hypokalemia although a supplementation with KCl-tablets may correct this disorder during PD therapy. Although the potassium losses through faeces, sweat and other body secretions were not measured in the present study and their contribution to serum potassium levels may be low, further studies are still needed to completely understand their roles in potassium balance in dialysis population. Our study indicated that, the dialysate prescription should be tailored to patients according to patients’ dietary intake, RRF as well peritoneal function status. It is interesting to note that in multiregression analysis, intracellular volume is the strongest predictor of serum potassium concentration in our patients. Although the accuracy of ECW and ICW measured by bioimpedance assessment is still an issue of debate, our results warrant further speculations. It is well known that although in PD the removed potassium is diffused from the extracellular space, potassium in the extracellular space is only a small part of the potassium pool in the body.20 Potassium is mainly present in the intracellular space where potassium concentration is about 30 times higher than that of potassium in the extracellular space.21 The distribution of potassium between the intracellular and extracellular fluid compartments is determined by many factors. The potassium shift through the cellular membrane follows passive diffusion from the intracellular to the extracellular compartment and, in the opposite direction, the active transport due to the Na–K ATPase transporter (Na–K pump).22 Factors that regulate Na–K pump in the active transport include catecholamine and insulin,23 while in the passive transport, potassium shift is regulated by the potassium concentration gradient between the intra- and the extracellular compartments. In fact, despite of removal via dialysis, insulin hormone, stimulated by the continuous glucose peritoneal infusion, could activate Na–K pump and generate an excess of potassium redistribution from extracellular into the intracellular compartment, leading to the decrease of plasma potassium.18 Moreover, acidosis could also reduce potassium intake by cells. There is a strong inverse relationship between blood pH and serum potassium concentration (SK), such that SK increases by about 0.8 mmol/L for each decline of 0.1 pH units.24 Subsequent equilibration
753
between the intra- and the extracellular compartments is driven by the potassium concentration gradient. Given the amount of potassium, the potassium concentration in intracellular compartment is determined by the distribution volume, ICW. It is thus not surprising to see that ICW significantly affects plasma potassium. Therefore, potassium removal in PD patients may not necessarily reflect potassium intake even if serum potassium is normal. In summary, our study suggested RRF plays an important role in potassium removal in CAPD patients. However, the regulation capacity in potassium metabolism of RRF was significantly impaired. Besides, the potassium level of CAPD patients also affected by intracellular volume, dietary intake, and dialysis dose. The appropriate PD dose should be given after detailed evaluation to avoid hypokalemia.
Acknowledgments The authors thank all the staff in the Division of Nephrology, Peking University Third Hospital.
Declaration of interest This study was partly supported by a grant from Beijing Municipal Science and Technology Foundation (D131100004713002) The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Khan AN, Bernardini J, Johnston JR, et al. Hypokalemia in peritoneal dialysis patients. Perit Dial Int. 1996;16(6):652. 2. Spital A, Sterns RH. Potassium supplementation via the dialysate in continuous ambulatory peritoneal dialysis. Am J Kidney Dis. 1985; 6(3):173–176. 3. Rostand SG. Profound hypokalemia in continuous ambulatory peritoneal dialysis. Arch Intern Med. 1983;143(2):377–378. 4. Burger CM. Emergency: hypokalemia. Am J Nurs. 2004; 104(11):61–65. 5. Szeto CC, Chow KM, Kwan BC, et al. Hypokalemia in Chinese peritoneal dialysis patients: prevalence and prognostic implication. Am J Kidney Dis. 2005;46(1):128–135. 6. Chuang YW, Shu KH, Yu TM, et al. Hypokalaemia: an independent risk factor of Enterobacteriaceae peritonitis in CAPD patients. Nephrol Dial Transplant. 2009;24(5):1603–1608. 7. Rippe B, Venturoli D. Optimum electrolyte composition of a dialysis solution. Perit Dial Int. 2008;28(3):131–136. 8. Alcazar AR. Electrolyte and acid-base balance disorders in advanced chronic kidney disease. Nefrologia. 2008; 28(3):87–93. 9. Redaelli B, Bonoldi G, Di Filippo G, et al. Behaviour of potassium removal in different dialytic schedules. Nephrol Dial Transplant. 1998;13(6):35–38. 10. Lu XH, Su CY, Sun LH, et al. Implementing continuous quality improvement process in potassium management in peritoneal dialysis patients. J Ren Nutr. 2009;19(6):469–474. 11. Nolph KD, Moore HL, Twardowski ZJ, et al. Cross-sectional assessment of weekly urea and creatinine clearances in patients on continuous ambulatory peritoneal dialysis. ASAIO J. 1992; 38(3):139–142. 12. Luo YJ, Lu XH, Woods F, et al. Volume control in peritoneal dialysis patients guided by bioimpedance spectroscopy assessment. Blood Purif. 2011;31(4):296–302.
754
H.-L. Yu et al.
Ren Fail Downloaded from informahealthcare.com by Biblioteka Uniwersytetu Warszawskiego on 01/12/15 For personal use only.
13. Enia G, Sicuso C, Alati G, et al. Subjective global assessment of nutrition in dialysis patients. Nephrol Dial Transplant. 1993; 8(10):1094–1098. 14. Rachoin JS, Weisberg LS. How should dialysis fluid be individualized for the chronic hemodialysis patient? Semin Dial. 2008; 21(3):223–225. 15. Beto JA, Bansal VK. Medical nutrition therapy in chronic kidney failure: integrating clinical practice guidelines. J Am Diet Assoc. 2004;104(3):404–409. 16. Kwok TC, Chan TY, Woo J. Relationship of urinary sodium/ potassium excretion and calcium intake to blood pressure and prevalence of hypertension among older Chinese vegetarians. Eur J Clin Nutr. 2003;57(2):299–304. 17. Chow KM, Szeto CC, Wong TY, et al. Risk factors for thiazideinduced hyponatraemia. QJM 2003;96(12):911–917. 18. Tziviskou E, Musso C, Bellizzi V, et al. Prevalence and pathogenesis of hypokalemia in patients on chronic peritoneal dialysis: one
Ren Fail, 2014; 36(5): 748–754
19. 20. 21. 22. 23. 24.
center’s experience and review of the literature. Int Urol Nephrol. 2003;35(3):429–434. Musso CG. Potassium metabolism in patients with chronic kidney disease. Part II: patients on dialysis (stage 5). Int Urol Nephrol. 2004;36(3):469–472. Edelman IS, Leibman J. Anatomy of body water and electrolytes. Am J Med. 1959;27:256–277. Ahmed J, Weisberg LS. Hyperkalemia in dialysis patients. Semin Dial. 2001;14(5):348–356. Ciandrini A, Severi S, Cavalcanti S, et al. Model-based analysis of potassium removal during hemodialysis. Artif Organs. 2009; 33(10):835–843. Clausen T, Everts ME. Regulation of the Na, K-pump in skeletal muscle. Kidney Int. 1989;35(1):1–13. Magner PO, Robinson L, Halperin RM, et al. The plasma potassium concentration in metabolic acidosis: a re-evaluation. Am J Kidney Dis. 1988;11(3):220–224.
