Renin-Angiotensin System Blockade Is Not Associated with Hyperkalemia in Chronic Hemodialysis Patients

Abstract

Background. Serious hyperkalemia was reported in 10% of chronic hemodialysis (HD) patients that could lead to arrhythmia and death. Angiotensin-converting enzyme inhibitors (ACE-I) and angiotensin II receptor blockers (ARB) are well accepted for cardio-protective benefits. The relationship between renin-angiotensin system blockade (RASB) and hyperkalemia in chronic HD patients remains controversial. The aim of this study was to find the relationship between RASB and hyperkalemia in these patients. Methods. Pre-dialysis serum potassium, clinical factors, and drugs were evaluated in 200 chronic HD patients in one HD center. Hyperkalemia was defined as serum K ≥ 5.3 meq/L. Finally, multivariate analysis with logistic regression was used to evaluate the risk of hyperkalemia by RASB and other factors. Results. In 200 patients, the mean K was 4.93 ± 0.79 meq/L, and 70 (35%) patients had hyperkalemia. Fifty-eight (29%) patients were prescribed with RASB. Seven variables—non-DM, longer HD duration, lower dialysate calcium, lower serum glucose, higher serum iPTH, not using RASB, and not using furosemide—were more frequent in hyperkalemia group. In logistic regression analysis, RASB was associated with decreased odds for hyperkalemia (OR 0.262, p = 0.001 in model A; OR 0.205, p = 0.001 in model B). In addition, furosemide was associated with decreased odds for hyperkalemia (OR 0.068, p = 0.022 in model B). Conclusions. RASB is not associated with hyperkalemia in chronic HD patients.

Keywords:

• RASB
• ACE-I
• ARB
• hyperkalemia
• hemodialysis

INTRODUCTION

Serious hyperkalemia was reported in 10% of chronic hemodialysis (HD) patients, which could lead to arrhythmia and death.^[1–3] Cardiovascular disease is the leading cause of mortality in chronic dialysis patients. Angiotensin-converting enzyme inhibitor (ACE-I) and angiotensin II receptor blocker (ARB) are well accepted for the cardio-protective benefits. Renin-angiotensin system blockade (ACE-I or ARB) is well known to be associated with hyperkalemia in patients with advanced chronic kidney disease (CKD) who are not on dialysis. However, few studies have investigated the relationship between renin-angiotensin system blockade (RASB) and hyperkalemia in chronic HD patients, and the results remain controversial.^[4,5]

In general, many factors, including dietary intake, DM, urine output, stool passage, KT/V, serum HCO₃⁻ level, and drugs such as RASB, heparin, furosemide, β-blocker, NSAID, and insulin may be related to hyperkalemia in chronic HD patients.^[6]

To examine the relationship between RASB and hyperkalemia in the present clinical practice for chronic HD patients, we conducted the observational study in 200 chronic HD patients.

SUBJECTS AND METHODS

Patient Selection

Two hundred patients receiving chronic HD (≥ 3 months, three times per week; ≥4.0 hours per session; blood flow ≥200 cc/min) in China Medical University Hospital (CMUH) were evaluated. Patients prescribed with potassium-lowering resin (Kayexalate), gastroenterol bleeding, and blood transfusion or hospitalization within two weeks was excluded. Informed consent was obtained. The study was approved by our ethics committee.

Study Design

Ours was a cross-sectional observational study of 200 chronic HD patients in June 2008 to examine the risk of hyperkalemia by RASB. Patients who entered the study continued all present medications and were re-educated with regular low potassium diet one week before blood sampling. The potassium concentration in dialysate was 2.0 meq/L in all patients. Blood samples were drawn one time from the arterial end of the vascular access before starting HD on Wednesday (for patients on a Monday-Wednesday-Friday hemodialysis schedule) or Thursday (for patients on a Tuesday-Thursday-Saturday hemodialysis schedule). Hyperkalemia was defined as pre-hemodialysis serum K ≥ 5.3 meq/L. Medication use was determined by chart record and confirmed by hemodialysis nurse. One hemodialysis nurse helped patients to complete the questionnaire of stool passage (classified as < 4 times per week or ≥4 times per week) and urine output (classified as <100 cc per day or ≥ 100 cc per day). Variables (i.e., age, gender, DM, urine output, stool passage, HD duration, HD modality, anticoagulants, dialysate glucose, dialysate calcium concentration, KT/V, serum HCO₃⁻, glucose, intact parathyroid hormone [iPTH] level, and drugs including ACE-I or ARB, furosemide, β-blocker, calcium channel blocker [CCB], digoxin, NSAID, aspirin, insulin, oral hypoglycemic agent [OHA], and Chinese herb) between hyperkalemia and non-hyperkalemia groups were evaluated.

Statistical Analysis

Statistical analysis was performed using SPSS 12.0 for Windows statistical software. Each value is expressed as the means ± SD. Continuous and ordinal data were analyzed using the Student t-test. Categorical data were analyzed by Fisher's exact test or chi-square test. Multivariate-adjusted odds ratios (ORs) for hyperkalemia and their 95% confidence intervals (CIs) were determined using logistic regression. Differences were considered significant if p < 0.05.

RESULTS

In a total of 200 patients, the mean serum K level was 4.93 ± 0.79 meq/L. Fifty-eight (29%) patients were prescribed with RASB. Seventy patients (35%) had hyperkalemia, defined as pre-hemodialysis K ≥ 5.3 meq/L (see Table 1). Patients with RASB had a lower mean K 4.68 ± 0.66 meq/L, compared with patients without RASB with a mean K 5.04 ± 0.81 meq/L, p = 0.03. Seven variables (viz., with non-DM, longer HD duration, lower dialysate calcium, higher serum iPTH, use of insulin, not using RASB, and not using furosemide) were significant in the hyperkalemia group (see Table 2).

Table 1  Patient characteristics, clinical data, and drugs

Age (year)	58.5 ± 14.3
Gender (male)	100 (50.0%)
DM	83 (41.5%)
Urine output (cc/day)
<100	137 (68.5%)
≥100	63 (31.5%)
Stool passage (times/week)
<4	58 (29%)
≥4	142 (71%)
HD duration (year)	4.91 ± 4.19
Dialysis modality
HD	191 (95.5%)
HDF	9 (4.5%)
Anticoagulants
None	8 (4.0%)
Low molecular weight heparin	33 (16.5%)
Heparin	159 (79.5%)
Dialysate glucose (mg/dl)
0	9 (4.5%)
100	143 (71.5%)
200	48 (24.0%)
Dialysate calcium (meq/l)
2.5	43 (21.5%)
3.0	109 (54.5%)
3.5	48 (24.0%)
KT/V	1.56 ± 0.29
Serum K (meq/l)	4.93 ± 0.79
<5.3	130 (65.0%)
≥5.3	70 (35.0%)
Serum HCO3^− (meq/l)	22.8 ± 2.7
Serum glucose (mg/dl)	109.5 ± 51.1
Serum iPTH (pg/ml)	196.4 ± 265.7
Drugs
RASB (ACE-I or ARB)	58 (29.0%)
Furosemide	13 (6.5%)
β-blocker	69 (34.5%)
Calcium channel blocker	97 (48.5%)
Digoxin	2 (1.0%)
NSAID	4 (2.0%)
Aspirin	42 (21.0%)
β-agonist	2 (1.0%)
Insulin	38 (19.0%)
Oral hypoglycemic agent	28 (14.0%)
Chinese herb	13 (6.5%)

Table 2  Characteristics of subjects by hyperkalemia

	K		p
Variable	<5.3 meq/l, N = 130	≥5.3 meq/l, N = 70
Age (year)	59.5 ± 15.3	56.7 ± 12.1	0.190
Gender (male)	65 (50%)	35 (50%)	1.000
DM	64 (49.2%)	19 (27.1%)	0.003
Urine output (<100 cc/day)	89 (68.5%)	48 (68.6%)	0.987
Stool passage (<4 times/week)	40 (30.1%)	18 (25.7%)	0.452
HD duration	4.3 ± 3.8	6.1 ± 4.6	0.003
Dialysis modality (HD)	126 (96.9%)	65 (92.9%)	0.282
Anticoagulant			0.356
None	7 (5.4%)	1 (1.4%)
Low molecular weight heparin	20 (15.4%)	13 (18.6%)
Heparin	103 (79.2%)	56 (80.0%)
Dialysate glucose (mg/dl)			0.238
0	6 (4.6%)	3 (4.3%)
100	88 (67.7%)	55 (78.6%)
200	36 (27.7%)	12 (17.1%)
Dialysate calcium (meq/l)			0.058
2.5	27 (20.8%)	16 (22.8%)
3.0	65 (50.0%)	44 (62.9%)
3.5	38 (29.2%)	10 (14.3%)
KT/V	1.6 ± 0.3	1.5 ± 0.2	0.362
Serum HCO3^− (meq/l)	22.9 ± 2.7	22.7 ± 2.5	0.661
Serum glucose (mg/dl)	114.4 ± 54.5	100.5 ± 42.8	0.066
Serum iPTH (pg/ml)	169.6 ± 266.8	246.0 ± 258.3	0.052
Drugs
RASB (ACE-I or ARB)	48 (36.9%)	10 (14.3%)	0.001
Furosemide	12 (9.2%)	1 (1.4%)	0.036
β-blocker	41 (31.5%)	28 (40.0%)	0.230
Calcium channel blocker	67 (51.5%)	30 (42.9%)	0.241
Digoxin	1 (0.8%)	1 (1.4%)	1.000
NSAID	3 (2.3%)	1 (1.4%)	1.000
Aspirin	32 (24.6%)	10 (14.3%)	0.087
β-agonist	1 (0.8%)	1 (1.4%)	1.000
Insulin	32 (24.6%)	6 (8.6%)	0.006
Oral hypoglycemic agent	20 (15.4%)	8 (11.4%)	0.442
Chinese herb	7 (5.4%)	6 (8.6%)	0.384

*p value < 0.05 or around 0.05.

In multivariate analysis of logistic regression, RASB was associated with decreased odds for hyperkalemia (OR 0.262, 95% CI 0.116–0.595, p = 0.001 in model A; OR 0.205, 95% CI 0.082–0.509, p = 0.001 in model B). In addition, furosemide was associated with decreased odds for hyperkalemia (OR 0.068, 95% CI 0.007–0.683, p = 0.022 in model B), as seen in Table 3.

Table 3  RASB associated with decreased odds for hyperkalemia after adjustment

Significant variable	Model A	Model B
RASB	OR 0.262	OR 0.205
	(95% CI 0.116–0.595) p = 0.001	(95% CI 0.082–0.509) p = 0.001
Furosemide	—	OR 0.068
		(95% CI 0.007–0.683) p = 0.022

Model A: adjusted by 6 significant covariates in table 2 (DM, HD duration, dialysate calcium, serum iPTH, furosemide, and insulin)

Model B: adjusted by covariates in model A plus other clinical factors, including age, gender, urine output, stool passage, KT/V, serum glucose, and HCO₃⁻

DISCUSSION

RASB is frequently indicated in renal failure patients. The relationship between RASB and hyperkalemia in patients with end-stage renal disease (ESRD) receiving chronic HD remains controversial. One study published in 2002 found that the use of ACE-I or ARB was independently associated with an increased risk of developing hyperkalemia in chronic HD patients.^[4] However, another one published in 2007 found that neither monotherapy (ACEI or ARB) nor combination therapy (ACEI plus ARB) was associated with the risk of hyperkaliemia in chronic HD patients.^[5] Our study also demonstrated that RASB (ACE-I or ARB) was not associated with hyperkalemia in chronic HD patients.

In our study, we defined hyperkalemia as pre-hemodialysis serum K ≥ 5.3 meq/L because of blood sampling by two-day interval (on Wednesday or Thursday) but not by three-day interval (on Monday or Tuesday). Fewer patients with RASB had hyperkalemia. Two reasons might account for this finding. One was that the prescribing physicians might be influenced by the knowledge that RASB could cause hyperkalemia in CKD patients who were not on dialysis and would not prescribe RASB to chronic HD patients with relatively higher baseline K level. The other was that RASB was really not associated with hyperkalemia in chronic HD patients.

In our results, the multivariate analysis of logistic regression after adjustment in different models really showed RASB was associated with decreased odds for hyperkalemia. Though some reports^[4,7,8] ever showed that RASB inhibited the residual renal excretion of potassium, colonic excretion of potassium, or cellular uptake of potassium, it might be possible that when ESRD patients received chronic HD to regularly remove body potassium, RASB was not associated with hyperkalemia in them. The finding of our study also showed furosemide with decreased odds for hyperkalemia. It suggested that chronic HD patients without furosemide might increase the risk of hyperkalemia.

The strength of this study was the consideration of the most possible factors causing hyperkalemia. However, there were several limitations of this study. First, although all patients were educated about a low potassium diet before blood sampling, some patients might not follow the diet accurately, and the potassium content of the patients' food were not measured. Second, although we classified patients according to stool passage, urine output, and use of furosemide or not, we did not measure the amount of potassium excretion in the urine or stool. Third, we could not exclude the selection bias that physicians might not prescribe RASB to patients with previous relatively higher baseline K level and the observational nature of this study.

CONCLUSION

Our findings show that renin-angiotensin system blockade is not associated with hyperkalemia in chronic hemodialysis patients.

ACKNOWLEDGMENTS

This study was supported by the China Medical University Hospital, DMR-97-037.