Kidney International, Vol. 61 (2002), pp.

21572164

CLINICAL NEPHROLOGY EPIDEMIOLOGY CLINICAL TRIALS

Calcium channel blocker use and mortality among patients

with end-stage renal disease
BRYAN KESTENBAUM, DANIEL L. GILLEN, DONALD J. SHERRARD, STEVEN SELIGER,
ADRIANNE BALL, and CATHERINE STEHMAN-BREEN
Division of Nephrology, School of Public Health and Community Medicine, and Department of Biostatistics, University of
Washington; and Division of Nephrology, Puget Sound Health Care System, Seattle, Washington, USA

Calcium channel blocker use and mortality among patients

with end-stage renal disease.
Background. Patients on dialysis suffer from alarming rates
of cardiovascular disease. While calcium channel blockers
(CCBs) are prescribed widely to patients with end-stage renal
disease (ESRD) for the treatment of hypertension, the longterm outcomes associated with the use of these medications
are not known. We sought to determine the association between CCB use and mortality among a cohort of ESRD patients.
Methods. Data were utilized from the United States Renal
Data System Dialysis Morbidity and Mortality Wave II, a randomly selected prospective cohort of 4065 ESRD patients who
began dialysis in 1996. Clinical data, including medication information, were collected 60 days after the start of dialysis. Subsequent survival status and cause of death were ascertained.
The Cox proportional hazards model was used to estimate the
relative risk of death associated with CCB use.
Results. Data from 3716 patients (91.4%) were available for
analysis. Fifty-one percent of the study patients were prescribed
a CCB. The use of a CCB was associated with a 21% lower
risk of total mortality (RR 0.79, CI 0.69 to 0.90) and a 26%
lower risk of cardiovascular specific mortality (RR 0.74, CI 0.60
to 0.91). For patients with pre-existing cardiovascular disease,
CCB use was associated with a 23% (RR 0.77, CI 0.65 to 0.91)
and 32% (RR 0.68, CI 0.53 to 0.87) lower risk of total and
cardiovascular mortality, respectively.
Conclusion. After controlling for known risk factors and
potential confounders, CCBs were found to be associated with
a lower risk of mortality among ESRD patients.

The risk of cardiovascular disease among patients with

end-stage renal disease (ESRD) is 10 times higher than
the general population [1]. While cardiovascular disease
remains the leading cause of death and disability for
patients with renal failure, few studies have investigated

whether cardiac medications can improve clinical outcomes for dialysis patients.
Among the general population, studies examining calcium channel blocker (CCB) use have reached mixed
conclusions with regard to their effects on patient outcome [28]. Short acting dihydropyridines have been
associated with an increased risk of myocardial infarction, while longer acting CCBs hold mortality risks
similar to other antihypertensive medications. Although
calcium channel blockers are prescribed widely to patients with ESRD, principally for the treatment of hypertension, the impact of CCB use on survival has never
been evaluated in this population. CCBs may have different effects in ESRD patients. Chronic renal failure
(CRF) generates a unique biochemical milieu characterized by derangements in calcium metabolism. In the setting of renal failure, calcium channel blockers reverse
pathologic levels of intracellular calcium that may contribute to the development of accelerated cardiovascular
disease [911]. Therefore, generalization of results from
non-uremic patients to those with renal failure should
be done with caution.
Based on the experimental benefit of CCBs in uremia,
we hypothesized that these medications would be beneficial to patients with renal failure. We conducted a cohort study using data from the United States Renal Database System Dialysis Morbidity and Mortality Study
Wave II (USRDS DMMS II) to estimate the risk of
death associated with CCB use among ESRD patients.
METHODS
Patient population

Key words: calcium channel blocker, dialysis, mortality, cardiovascular

disease, hypertension, dihydropyridines, USRDS DMMS II.
Received for publication May 24, 2001
and in revised form January 10, 2002
Accepted for publication January 11, 2002

2002 by the International Society of Nephrology

Details of the USRDS DMMS II are described in

detail elsewhere [12]. Briefly, the USRDS collects demographic and clinical data on all patients who have survived more than 90 days on dialysis. DMMS II was a
prospective study containing a random sample of 4065
incident hemodialysis (HD) and peritoneal dialysis (PD)

2157

2158

Kestenbaum et al: Calcium channel blockers and ESRD

patients initiating dialysis in 1996 and early 1997 from

25% of United States dialysis facilities (N 799). To
obtain comparable numbers of PD and HD patients for
the study, PD patients were over sampled. All incident
PD patients were included whereas only 20% of all HD
patients were selected. Patients were excluded from
DMMS II if they were younger than 18 years, home
dialysis patients, or previously transplanted. For the purposes of this analysis, all subjects who participated in
DMMS II were included.
Data collection
Dialysis center personnel collected baseline and follow-up patient data using medical records from dialysis
centers, hospitalizations, and personal physicians, and
by directly interviewing patients. Dialysis modality was
determined on day 60 of ESRD. Baseline data were
collected 60 days after the start of dialysis. Variables
used for the purpose of this analysis were patient gender,
age, race (black, white, and other), smoking status
(never, former, or current smoker), treatment modality
(hemodialysis, peritoneal dialysis, transplant), medications, serum albumin, pre-dialysis blood pressure (systolic and diastolic), history of diabetes (yes, no), clinical
determination of malnourishment (yes, no), and history
of the following specific medical conditions: coronary
artery disease, myocardial infarction, coronary angioplasty, coronary artery bypass surgery, cardiac arrest,
and stroke. These clinical conditions, including cardiac
history, were determined by dialysis personnel through
review of notation in the patients medical charts and
through interviews with the patients themselves. A maximum of 15 medications per patient were recorded. We
determined if CCBs, angiotensin-converting enzyme
(ACE) inhibitors, blockers and aspirin had been prescribed by comparing patients recorded medications to
a list of known generic and brand names. CCBs were
classified into four groups: group 1, short acting nifedipine; group 2, long acting dihydropyridines; group 3,
diltiazem; and group 4, verapamil. Changes in treatment
modality during follow-up were obtained from the Treatment History File, which is maintained by the USRDS.
Ascertainment of outcome
Survival status and cause of death were linked to the
Wave 2 data from the USRDS Patients Standard Analysis File (SAF) via unique patient identifiers assigned
by the USRDS. Patient survival status is periodically
updated in the SAF and at the time of analysis was
complete through July 1998. The date and cause of death
listed in the SAF were obtained by the USRDS from
HCFA form #2746, which is completed by the primary
nephrologist following the death of any dialysis patient.
For the purposes of this analysis, death from cardiovascular disease was defined a priori as death from myocardial

infarction, atherosclerotic heart disease, cardiomyopathy, cerebrovascular accident, cardiac arrhythmia and
cardiac arrest of unknown cause.
Statistical analysis
The association between baseline CCB use and time to
death was analyzed using the Cox proportional hazards
model for censored failure times. In this model, the hazard or the instantaneous probability of death, was modeled as a function of the predictor covariates. The relative
risk (RR) or hazard ratio was then estimated for each
covariate as the proportionate change in the instantaneous probability of death for two individuals differing
only by a single unit of that covariate. A relative risk less
than one suggests that a one-unit increase in a covariate is
associated with a longer time to death. Alternatively, a
relative risk greater than one suggests a shorter time to
death. Variables used in the multivariate model were
chosen a priori and retained in the model if there was
biological plausibility or if exploratory analyses suggested that the covariate of interest may be associated
with death or may confound the relationship between
CCB use and death. Variables used for adjustment in
the models included age, sex, race, treatment modality,
diabetes, pre-existing cardiovascular disease, undernourishment, serum albumin, pre-dialysis systolic and diastolic blood pressure, and the use of an ACE inhibitor,
blocker, and aspirin. Formal tests as well as graphical
methods were used to verify the existence of proportional hazards. Further, residual diagnostics were used
to identify outlying points and to model the correct functional form of adjustment variables. Estimated RR along
with corresponding 95% confidence intervals and P values for two-sided tests of association are reported for
all regression covariates.
To improve generalization of our results, patients were
not censored at the time of transplant. Since transplant
patients are generally healthier, censoring patients at the
time of transplant would have made our results applicable only to less healthy ESRD patients. Because the
dialysis modality administered to a patient can change
over time, treatment modality was entered into all analyses as a time dependent covariate. The use of the timedependent covariate allowed patients to continue to contribute time at risk for a given modality for the amount
of time they underwent that particular modality.
RESULTS
A total of 4065 patients were included in DMMS II.
Of these, 349 patients were excluded from our analysis
because they did not receive a valid USRDS patient
identifier, making follow-up impossible. Therefore, 3716
patients were available for analysis. There were no notable differences in clinical characteristics between patients
included and excluded from the study.

2159

Kestenbaum et al: Calcium channel blockers and ESRD

Table 1. Patient characteristics at the time of data collection

Characteristic
Age years
Female sex
Race
White
Black
Other
Smoking status
Never
Former
Current
History of cardiovascular diseasea
Undernourished
History of diabetes
Mean serum albumin mg/dL
Mean serum cholesterol
Mean serum triglycerides
Mean pre-dialysis diastolic BP mm Hg
Mean pre-dialysis systolic BP mm Hg
Aspirin use (yes vs. no)
ACE inhibitor use (yes vs. no)
blocker use (yes vs. no)

No calcium
channel
blocker use
(N 1814)

Calcium
channel
blocker use
(N 1902)

59.9 (15.7)
857 (47.2%)

58.2 (15.8)
890 (46.8%)

1211 (67.5%) 1147 (60.8%)

450 (25.1%) 567 (30.0%)
134 (7.5%)
173 (9.2%)
960
509
196
795
376
874
3.42
188.6
190.5
78.5
141.7
247
413
289

(57.7%) 991 (55.5%)

(28.5%) 510 (30.5%)
(11.8%) 286 (16.0%)
(49.4%) 833 (48.3%)
(21.6%) 326 (17.7%)
(49.1%) 951 (50.9%)
(0.58)
3.50 (0.58)
(54.7)
198.4 (56.7)
(145.3) 206.5 (142.1)
(11.9)
81.8 (11.9)
(21.1)
150.9 (19.0)
(13.6%) 317 (16.7%)
(22.8%) 446 (23.4%)
(15.9%) 372 (19.6%)

Data are mean (SD) or N (%)

a
Defined as a prior diagnosis of coronary heart disease, coronary artery disease, myocardial infarction, coronary artery bypass surgery, angioplasty, cardiac
arrest, cerebrovascular accident, or congestive heart failure

The baseline characteristics of the study patients are

shown in Table 1. Fifty-one percent of our cohort reported taking a CCB 60 days after the start of dialysis.
There were no important differences in baseline clinical
characteristics between patients who reported using a
CCB, and those who did not.
Table 2 describes the proportion of patients using
CCBs stratified by CCB group. The majority of subjects
prescribed CCBs were receiving long acting dihydropyridines, particularly amlodipine and long acting nifedipine.
Diltiazem was prescribed to approximately 10% of the
cohort, while verapamil was rarely used. A few patients
(1.4%) reported using a combination of two CCBs. The
most frequent combinations reported were a dihydropyridine plus diltiazem, followed by a dihydropyridine plus
short acting nifedipine.
There were 1232 total deaths during the follow-up
period. Table 3 presents adjusted and unadjusted relative
risks of total mortality associated with the covariates of
interest. Among CCB users, the total mortality rate was
169.8 deaths/1000 person-years compared to 225.6 deaths/
1000 person-years among non-users. The use of a betablocker, aspirin, or an ACE inhibitor was not associated
with a difference in the adjusted risk of death from all
causes, while the use of a CCB was significantly associated with a lower risk of total mortality. For patients
reporting the use of any CCB, the adjusted relative risk
for total mortality was 0.79 (CI 0.69 to 0.90) when compared to patients that did not report CCB use, while for

Table 2. Prevalence of calcium channel blocker use (N 3716)

Calcium channel blocker class
Any calcium channel blocker
Class I
Nifedipine (Procardia)
Class II
Nifedipine (Procardia, Adalat) XL, CC, or GITS
Amlodipine (Norvasc)
Felodipine (Plendil)
Isradipine (Dynacirc)
Nicardipine (Cardene)
Nisoldipine
Lacidipine
Class III
Diltiazem (Cardizem, Dilacor)
Class IV
Verapamil (Isoptin)
Combination of 2 classes

N (%)
1902
351
351
1162
449
637
40
35
10
0
0
363
363
76
76
50

(51.2%)
(9.5%)
(9.5%)
(31.3%)
(12.1%)
(17.1%)
(1.1%)
(0.9%)
(0.3%)
(0%)
( 0%)
(9.8%)
(9.8%)
(2.0%)
(2.0%)
(1.4%)

patients reporting diltiazem use the adjusted relative risk

was estimated to be 0.63 (CI 0.49 to 0.81). Of the clinical
characteristics we studied, advancing age, low serum albumin, diabetes, pre-existing cardiac disease, and Caucasian race were all significantly associated with higher
risks of total mortality. Blood pressure had a U-shaped
relationship with mortality; pre-dialysis systolic pressures less than 130 and greater than 160 were associated
with a higher risk of death. Similarly, pre-dialysis diastolic pressures less than 60 and greater than 90 were
associated with a higher risk of mortality. We tested several interaction terms, including an interaction between
CCB use and treatment modality, and an interaction
between CCB use and race. These interaction terms were
found not to be statistically significant.
Table 4 lists the adjusted and unadjusted relative risks
for cardiovascular specific mortality associated with the
identified baseline characteristics. There were 72.45 cardiovascular related deaths/1000 person-years among
CCB users as compared to 107.75 CV related deaths/
1000 person-years among non-users. After adjustment,
the use of a blocker, ACE inhibitor, or aspirin was
not associated with a difference in cardiovascular specific
mortality. In contrast, many of the calcium channel
blockers were associated with a significantly lower risk
of cardiovascular death. For example, the use of any
calcium channel blocker was associated with a 26% lower
risk of cardiovascular specific death (aRR 0.74, CI 0.60
to 0.91). Similarly, the use of the non-dihydropyridine,
diltiazem was associated with 38% lower risk of cardiovascular mortality (aRR 0.62, CI 0.42 to 0.90). We did
not observe a higher risk of CV mortality associated with
the prescription of short acting nifedipine. Pre-existing
cardiovascular disease, low serum albumin, undernourishment, and diabetes were associated with higher risks
of cardiovascular specific death.
Table 5 illustrates the results of our analysis after stratifying by previous history of cardiovascular disease. For

2160

Kestenbaum et al: Calcium channel blockers and ESRD

Table 3. Cox regression results for all-cause mortality

Covariate
Age per decade
Female sex
Race
White
Black
Other
Modality
HD
PD
Transplant
Smoking Status
Never
Former
Current
History of CVDa
Undernourished (yes vs. no)
History of diabetes
Albumin 1 g/dL decrease
Diastolic BP mm Hg
60 to 90
Less than 60
Greater than 90
Systolic BP mm Hg
130 to 160
Less than 130
Greater than 160
Aspirin (yes vs. no)
ACE inhibitors (yes vs. no)
Blockers (yes vs. no)
CCB use (yes vs. no)
No reported use
Class I
Class II
Class III
Class IV
Combination of 2
a

Deaths
(N 1232)

Unadjusted RR
(95% CI)

P value

Adjusted RR
(95% CI)

P value

1225
601

1.48 (1.42, 1.55)

1.09 (0.97, 1.21)

0.001
0.147

1.30 (1.23, 1.38)

1.12 (0.98, 1.29)

0.001
0.104

897
267
63

1.0
0.62 (0.54, 0.71)
0.48 (0.37, 0.62)

0.001
0.001

1.0
0.74 (0.63, 0.88)
0.58 (0.43, 0.78)

0.001
0.001

720
468
15

1.0
0.96 (0.86, 1.08)
0.17 (0.10, 0.28)

0.550
0.001

1.0
1.07 (0.92, 1.23)
0.29 (0.15, 0.57)

0.374
0.001

1.0
(1.13,
(0.80,
(2.45,
(1.93,
(1.36,
(1.67,

1.46)
1.15)
3.17)
2.48)
1.70)
2.00)

0.001
0.652
0.001
0.001
0.001
0.001

0.99
1.22
1.66
1.43
1.31
1.43

1.0
(0.85,
(0.99,
(1.42,
(1.23,
(1.14,
(1.26,

1.16)
1.51)
1.94)
1.67)
1.52)
1.62)

0.928
0.063
0.001
0.001
0.001
0.001

1.0
2.67 (2.23, 3.21)
0.55 (0.45, 0.66)

0.001
0.001

1.0
1.36 (1.07, 1.74)
0.98 (0.77, 1.24)

0.013
0.842

2.13)
1.28)
1.62)
1.08)
1.09)
0.83)

0.001
0.154
0.001
0.365
0.419
0.001

1.53
1.13
0.95
0.96
1.03
0.79

0.87)
0.88)
0.88)
1.14)
1.29)

0.001
0.001
0.002
0.163
0.327

0.73
0.89
0.63
0.48
0.65

608
387
142
750
360
712
1126
946
132
123
564
349
288
234
270
207
561
672
92
340
95
19
15

1.28
0.96
2.79
2.19
1.52
1.82

1.86
1.11
1.41
0.94
0.94
0.74
0.70
0.77
0.71
0.72
0.77

1.0
(1.63,
(0.96,
(1.22,
(0.82,
(0.81,
(0.67,
1.0
(0.56,
(0.68,
(0.57,
(0.46,
(0.46,

1.0
(1.28,
(0.95,
(0.79,
(0.82,
(0.87,
(0.69,
1.0
(0.56,
(0.76,
(0.49,
(0.24,
(0.36,

1.82)
1.33)
1.13)
1.13)
1.22)
0.90)

0.001
0.168
0.544
0.637
0.749
0.001

0.96)
1.04)
0.81)
0.97)
1.19)

0.022
0.157
0.001
0.040
0.165

Defined as prior Dx of CHD/CAD, MI, bypass, angioplasty, cardiac arrest, cerebrovascular accident or congestive heart failure

patients without a prior history of cardiovascular disease,

the use of a CCB was not associated with a significant
change in the risk of total or cardiovascular related death.
In contrast, a lower risk of mortality was found among
patients with pre-existing CVD who reported using a
calcium channel blocker. After adjustment, the use of
any CCB was associated with a 32% lower risk of death
from cardiovascular disease (aRR 0.68, CI 0.53 to 0.87)
and a 23% lower risk of death from all causes (aRR
0.77, CI 0.65 to 0.91). The lowest associated risk was
observed for the more cardiac selective calcium channel
blockers, diltiazem, and verapamil. For patients with a
prior history of cardiovascular disease, diltiazem was
associated with a 48% lower risk of cardiovascular specific death (aRR 0.52, CI 0.33 to 0.82) and 38% lower
risk of death from all causes (aRR 0.62, CI 0.46 to 0.84).
DISCUSSION
Greater than half of the ESRD patients under study
reported taking a calcium channel blocker. The use of
any CCB was associated with a 21% lower risk of all-

cause mortality and a 26% lower risk of cardiovascular

specific mortality. After stratifying by history of cardiovascular disease, no association was found between CCB
use and mortality for patients without a previous history
of CVD. In contrast, CCB use was associated with a
32% lower risk of cardiovascular mortality for patients
who reported a prior history of CVD.
There was a lower risk of total and cardiovascular
mortality associated with CCB use. While our study design is observational, these results are consistent with
those from randomized trials among the general population comparing CCBs to placebo for the treatment of
hypertension [5, 7]. For example, nitrendipine was found
to reduce total mortality by 55% versus placebo when
prescribed to 492 elderly, hypertensive diabetic patients
[7]. When felodipine was used as primary therapy to
treat 18,790 hypertensive patients in the HypertensionOptimal-Therapy trial, a trend toward decreased cardiovascular mortality was observed as blood pressures were
lowered [5]. We found a lower relative risk of mortality
associated with CCB use among our cohort than generally reported among patients without renal failure. It is

2161

Kestenbaum et al: Calcium channel blockers and ESRD

Table 4. Cox regression results for CV/stroke-related mortality
Covariate
Age per decade
Female sex
Race
White
Black
Other
Modality
HD
PD
Transplant
Smoking Status
Never
Former
Current
History of CVDa
Undernourished (yes vs. no)
History of diabetes
Albumin 1 g/dL decrease
Diastolic BP mm Hg
60 to 90
Less than 60
Greater than 90
Systolic BP mm Hg
130 to 160
Less than 130
Greater than 160
Aspirin (yes vs. no)
ACE inhibitors (yes vs. no)
Beta-blockers (yes vs. no)
CCB use (yes vs. no)
No reported use
Class I
Class II
Class III
Class IV
Combination of 2
a

Deaths
(N 560)

Unadjusted RR
95% CI

P value

Adjusted RR
95% CI

P value

555
256

1.56 (1.47, 1.67)

0.96 (0.81, 1.14)

0.001
0.650

1.35 (1.23, 1.47)

1.03 (0.84, 1.27)

0.001
0.756

409
117
31

1.0
0.60 (0.49, 0.74)
0.52 (0.36, 0.75)

0.001
0.001

1.0
0.78 (0.61, 1.01)
0.59 (0.38, 0.92)

0.055
0.021

321
229
3

1.0
1.03 (0.87, 1.22)
0.09 (0.03, 0.27)

0.745
0.001

1.0
1.24 (1.00, 1.53)
0.11 (0.02, 0.74)

0.047
0.023

279
185
51
295
146
342
515

1.33
0.73
3.54
1.87
1.73
1.64

1.0
(1.10,
(0.54,
(2.89,
(1.55,
(1.46,
(1.43,

1.60)
0.98)
4.33)
2.26)
2.06)
1.89)

0.003
0.035
0.001
0.001
0.001
0.001

0.97
0.93
1.89
1.20
1.59
1.22

1.21)
1.32)
2.41)
1.52)
1.97)
1.48)

0.791
0.682
0.001
0.138
0.001
0.046

433
69
45

1.0
2.96 (2.29, 3.81)
0.44 (0.33, 0.60)

0.001
0.001

1.0
1.64 (1.17, 2.31)
0.83 (0.56, 1.22)

0.004
0.335

2.44)
1.50)
2.08)
1.13)
1.21)
0.79)

0.001
0.064
0.001
0.431
0.804
0.001

1.58
1.32
1.13
0.90
1.02
0.74

0.87)
0.83)
0.97)
0.97)
1.62)

0.006
0.001
0.029
0.042
0.486

0.75
0.80
0.62
0.27
0.82

245
165
137
125
121
97
239
321
39
143
45
5
7

2.00
1.22
1.71
0.92
0.97
0.67
0.63
0.68
0.71
0.40
0.77

1.0
(1.64,
(0.99,
(1.40,
(0.75,
(0.78,
(0.57,
1.0
(0.45,
(0.56,
(0.52,
(0.17,
(0.36,

1.0
(0.78,
(0.65,
(1.49,
(0.94,
(1.28,
(1.00,

1.0
(1.21,
(1.03,
(0.88,
(0.71,
(0.79,
(0.60,
1.0
(0.51,
(0.63,
(0.42,
(0.07,
(0.36,

2.06)
1.69)
1.44)
1.14)
1.31)
0.91)

0.001
0.030
0.342
0.391
0.876
0.004

1.11)
1.02)
0.90)
1.07)
1.85)

0.155
0.07
0.013
0.063
0.628

Defined as prior Dx of CHD/CAD, MI, bypass, angioplasty, cardiac arrest, cerebrovascular accident, or congestive heart failure

Table 5. Cox regression results stratified by history of CVD

Covariate
CV-related mortality
CCB use (yes vs. no)
No reported use
Class I
Class II
Class III
Class IV
Combination of 2
All-cause mortality
CCB use (yes vs. no)
No reported use
Class I
Class II
Class III
Class IV
Combination of 2

No prior history
aRRa (95% CI)
0.86 (0.58,
1.0
1.04 (0.53,
0.79 (0.50,
0.98 (0.49,
0.44 (0.06,
1.06 (0.14,
0.79 (0.62,
1.0
0.96 (0.63,
0.81 (0.61,
0.65 (0.39,
0.68 (0.25,
0.60 (0.15,

P value

1.27)

0.434

2.05)
1.25)
1.93)
3.19)
7.81)

0.901
0.316
0.944
0.413
0.953

1.01)

0.062

1.47)
1.07)
1.06)
1.86)
2.46)

0.856
0.137
0.082
0.451
0.480

Prior history
aRRa (95% CI)
0.68 (0.53,
1.0
0.65 (0.40,
0.78 (0.59,
0.52 (0.33,
0.19 (0.03,
0.68 (0.28,
0.77 (0.65,
1.0
0.64 (0.45,
0.90 (0.75,
0.62 (0.46,
0.37 (0.14,
0.62 (0.32,

P value

0.87)

0.002

1.06)
1.03)
0.82)
1.38)
1.67)

0.085
0.078
0.005
0.101
0.397

0.91)

0.002

0.90)
1.09)
0.84)
1.00)
1.21)

0.011
0.299
0.002
0.050
0.161

a
Adjusted for age, sex, race, treatment modality, smoking status, history of CVD, undernourished, albumin, diastolic BP, systolic BP, aspirin use, ACE inhibitor
use and beta blocker use

possible that the actions of CCBs, such as reduction in

blood pressure, attenuation of left ventricular hypertrophy (LVH), and restoration of intracellular calcium are
of particular benefit to patients with ESRD.

In contrast to some reports, which have suggested

that CCBs increase the risk of myocardial infarction and
death among selected patient groups when compared
to other antihypertensive agents [3, 6, 8], there was no

2162

Kestenbaum et al: Calcium channel blockers and ESRD

association between the use of a CCB and a greater risk

of mortality among ESRD patients. Recently published
randomized trials, such as the International Nifedipine
GITS Study and the Nordic Diltiazem Study, have found
similar mortality rates among users of CCBs, blockers
and diuretics [2, 4]. The more gradual reduction in blood
pressure produced by newer, long acting forms of calcium channel blockers may account for their enhanced
safety.
We found a particularly strong association between
CCB use and lower mortality among patients with prior
cardiovascular disease (CVD). Studies from the general
population have reached differing conclusions regarding
CCB therapy among patients who have suffered acute
MI [8, 1317]. The Multi-Center Diltiazem Post Infarction Trial (MDPIT) found no difference in survival
for 2466 patients assigned diltiazem or placebo following
MI [13]. However, subsets of MDPIT subjects, such as
those without pulmonary congestion at presentation and
those with hypertension, experienced a reduction in cardiac death or re-infarction with diltiazem use [18]. The
Danish Verapamil Infarction Trial II (DAVIT II), examining verapamil therapy following acute MI, found a
non-significant 20% reduction in death or re-infarction
for patients randomized to verapamil versus placebo
[14]. When similar patients from MDPIT and DAVIT II,
with non-Q wave MI and without pulmonary congestion,
were combined, the relative risk of mortality for users
of diltiazem and verapamil was 0.65 (CI 0.40 to 1.05)
[15], similar to our estimates among ESRD patients.
There were no statistically significant associations between CCB use and mortality among patients without
prior CVD in our cohort. It is possible that CCBs are
of particular benefit for ESRD patients with CVD. Alternatively, we may have lacked the statistical power to
detect significant associations among patients without
prior CVD because these patients experienced fewer
events, though the estimated relative risks associated
with CCB use suggests the former.
No significant interaction was found between CCB
use, race, and mortality, implying that the lower risk of
death associated with CCB use was similar among white
and African American patients under this study. Among
African American patients with hypertension and renal
insufficiency participating in the African American Study
of Kidney Disease and Hypertension (AASK) trial, the
risk of death was not statistically different among subjects treated with CCBs as compared to ACE inhibitors
[19]. However, the population studied and end points
used for the AASK trial were different from the present
study, limiting direct comparisons.
While most studies have found long acting CCBs to
be safe for the treatment of hypertension, short acting
nifedipine was reported to increase the risk of MI for
patients with hypertension, as well as for patients with

a recent MI [3, 20]. In contrast, our results show a lower

mortality for ESRD patients who reported taking short
acting nifedipine. It is possible that the results of previous
studies do not apply directly to our cohort, either because
of differences in design or because of differences in the
response of ESRD patients to this medication. Additionally, it is possible that complete medication names, such
as XL, GITS, or CC were not accurately reported in the
USRDS, resulting in misclassification and reducing our
ability to detect changes in mortality associated with
short acting nifedipine use.
Our data do not show an association between the use
of ACE inhibitors, blockers, or aspirin and the risk
of mortality among ESRD patients. Patients with renal
failure have unique physiologic characteristics, which
may lead to unpredictable responses to traditional cardiac medications. For example, low renin levels found
in many dialysis patients could attenuate the benefit from
ACE inhibition and explain the lack of association observed between ACE inhibitor use and mortality among
our cohort. Similarly, the significant platelet dysfunction
induced by renal failure could be responsible for the
lack of association detected between aspirin use and
mortality. Finally, a blunted response to sympathetic
stimulation might explain a lack of association between
beta-blocker use and mortality among patients with renal
failure. Further exploration of this complex issue is necessary to optimize antihypertensive therapy among patients with ESRD.
Calcium channel blockers possess a variety of potential therapeutic properties in dialysis patients. Traditional mechanisms by which CCBs are believed to act
include relaxation of vascular smooth muscle, control
of blood pressure, and attenuation of heart rate and
contractility. These properties may be of special importance to dialysis patients, who have strikingly high rates
of hypertension and LVH [21, 22]. For example, nifedipine suppresses surges in fibrinogen and von Willebrand
factor that are produced by erythropoietin therapy [23].
In uremic animals with hypertension, CCBs increase stable nitric oxide (NOS) metabolites and enhance vascular
NOS activity [24]. Among the general population, CCBs
reduce the pathologic changes associated with LVH
[25, 26]. In one study of chronic dialysis patients, diltiazem treatment improved the hemodynamics of LVH, as
measured by echocardiography [27].
Alternatively, CCBs may be producing beneficial effects via their specific actions on intracellular calcium
(iCa) in the setting of renal failure. Pathologic iCa levels
in renal failure have been documented in a variety of cell
types, including neutrophils, pancreatic islets, platelets,
endothelial cells, and cardiac myocytes [10, 2830]. Accompanying these changes in iCa are significant abnormalities in cell function, including an inability to respond
to signals that use calcium as an intracellular messenger.

Kestenbaum et al: Calcium channel blockers and ESRD

For example, in myocytes, a rise in intracellular calcium

is associated with impaired response to IGF-1, and decreased protein synthesis [29]. The derangement of iCa
produced by renal failure has been linked to elevated
levels of circulating parathyroid hormone, and is normalized by parathyroidectomy, and by calcium channel
blockers [11, 31]. CCBs prevent an increase of intracellular calcium by directly antagonizing PTH at the cellular
level, blocking the influx of calcium stimulated by this
hormone [28, 3133]. It is possible that CCBs confer a
survival advantage among patients with ESRD by restoring levels of iCa.
Our study has several limitations. First, since CCB use
was assessed at a single point in time, we could not
determine the length of CCB treatment prior to data
collection or whether CCBs were in use at the time of
death. However, if users and non-users of CCBs were
equally likely to change medications without respect to
their subsequent mortality status, then non-selective misclassification of CCB use would occur. This type of misclassification would dilute the association between CCB
use and mortality leading to an underestimation of the
true relative risk. Non-selective misclassification may
have been a particular problem with aspirin use because
patients may not have considered aspirin to be a prescription medication. The result would be an underestimation
of the effect of aspirin use.
Although we controlled for many known factors that
influence mortality in ESRD patients, it is quite plausible
that other important factors were not measured. In order
to bias our results, these factors would have to disproportionately affect the users or non-users of CCBs and be
associated with mortality. It is possible that the indication
for a CCB per se identifies patients with inherent characteristics that influence their long-term mortality (confounding by indication). For example, some dialysis patients may experience low blood pressure or poor general
health, which could preclude them from being prescribed
a CCB. If a higher proportion of these patients were to
die compared to patients prescribed a CCB, we could
falsely conclude that CCBs have protective effects. However, in this case it is more likely that the intention to
prescribe a CCB identifies patients at a higher baseline
risk of death, because the common indications to prescribe a CCB are hypertension and atrial fibrillation,
which increase the risk of mortality. To further address
this problem, we included estimations of the relative
risks of death for other cardiovascular medications that
would be prescribed for similar reasons as CCBs within
our cohort. Furthermore, sensitivity analyses were performed and no significant change was found in our results
after adjustment for markers of pre-ESRD care, such as
erythropoietin use, the number of visits, and number of
months in which patients saw their nephrologist prior to
starting dialysis.

2163

Finally, it is possible that there was misclassification

of cause of death. A previous study by Perneger, Klag
and Whelton suggested that there is a poor correlation
between cause of death identified on the Death Notification Form and death certificates [34]. However, 40% of
deaths among patients with ESRD were classified as due
to renal failure on their death certificate. Since most
ESRD patients do not die of renal failure per se, the
results of Pernegers study suggests that the Death Notification Form may be more accurate than a death certificate. While our study has limitations, it is the first, to
our knowledge, to assess outcomes associated with CCB
use in the ESRD population. In addition, as a population
based study, the results are not a reflection of physicianor region-specific characteristics and therefore can be
generalized to the entire ESRD population.
In conclusion, the use of a calcium channel blocker is
associated with lower total and cardiovascular specific
mortality among a cohort of ESRD patients in the
USRDS DMMS II. This finding is particularly notable
for patients with a prior history of cardiovascular disease.
Further investigations are needed to confirm these results and help guide optimal therapy for this patient
population.
ACKNOWLEDGMENTS
This study was supported by a Department of Veterans Affairs
Career Development Award and a PHS grant (DK07721-6) from the
National Institutes of Health (Bethesda, MD, USA). The data reported
here have been supplied by the USRDS. The interpretation and reporting of these data are the responsibility of the authors and in no
way should be seen as an official policy of or interpretation by the
U.S. Government.
Reprint requests to Bryan Kestenbaum, M.D., Veterans Affairs Puget
Sound Health Care System, 1660 South Columbian Way, Mailstop 111A,
Seattle, Washington 98108, USA.
E-mail: epware@earthlink.net

REFERENCES
1. Foley RN, Parfrey PS, Sarnak MJ: Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 32
(Suppl):S112S119, 1998
2. Brown MJ, Palmer CR, Castaigne A, et al: Morbidity and mortality in patients randomised to double-blind treatment with a longacting calcium-channel blocker or diuretic in the International
Nifedipine GITS study: Intervention as a Goal in Hypertension
Treatment (INSIGHT). Lancet 356:366372, 2000
3. Psaty BM, Heckbert SR, Koepsell TD, et al: The risk of myocardial infarction associated with antihypertensive drug therapies.
JAMA 274:620625, 1995
4. Hansson L, Hedner T, Lund-Johansen P, et al: Randomised trial
of effects of calcium antagonists compared with diuretics and betablockers on cardiovascular morbidity and mortality in hypertension: The Nordic Diltiazem (NORDIL) study. Lancet 356:359
365, 2000
5. Hansson L, Zanchetti A, Carruthers SG, et al: Effects of intensive blood-pressure lowering and low-dose aspirin in patients with
hypertension: Principal results of the Hypertension Optimal Treatment (HOT) randomised trial. HOT Study Group. Lancet 351:
17551762, 1998
6. Estacio RO, Jeffers BW, Hiatt WR, et al: The effect of nisoldipine

2164

7.

8.

9.
10.
11.
12.
13.
14.
15.

16.
17.

18.

19.

Kestenbaum et al: Calcium channel blockers and ESRD

as compared with enalapril on cardiovascular outcomes in patients

with non-insulin-dependent diabetes and hypertension. N Engl J
Med 338:645652, 1998
Tuomilehto J, Rastenyte D, Birkenhager WH, et al: Effects
of calcium-channel blockade in older patients with diabetes and
systolic hypertension. Systolic Hypertension in Europe Trial Investigators. N Engl J Med 340:677684, 1999
Borhani NO, Mercuri M, Borhani PA, et al: Final outcome
results of the Multicenter Isradipine Diuretic Atherosclerosis
Study (MIDAS). A randomized controlled trial. JAMA 276:785
791, 1996
Zhang YB, Smogorzewski M, Ni Z, et al: Altered cytosolic calcium
homeostasis in rat cardiac myocytes in CRF. Kidney Int 45:1113
1119, 1994
Alexiewicz JM, Smogorzewski M, Akmal M, et al: Nifedipine
reverses the abnormalities in [Ca2]i and proliferation of B cells
from dialysis patients. Kidney Int 50:12491254, 1996
Horl WH, Haag-Weber M, Mai B, et al: Verapamil reverses
abnormal [Ca2]i and carbohydrate metabolism of PMNL of dialysis patients. Kidney Int 47:17411745, 1995
System USRDS: Researchers Guide to the USRDS Database.
Bethesda, The National Institutes of Health, National Institute of
Diabetes and Digestive and Kidney Diseases, 1999
The Multicenter Diltiazem Postinfarction Trial Research
Group: The effect of diltiazem on mortality and reinfarction after
myocardial infarction. N Engl J Med 319:385392, 1988
Effect of verapamil on mortality and major events after acute
myocardial infarction: The Danish Verapamil Infarction Trial II
DAVIT II. Am J Cardiol 66:779785, 1990
Gibson RS, Hansen JF, Messerli F, et al: Long-term effects of
diltiazem and verapamil on mortality and cardiac events in nonQ-wave acute myocardial infarction without pulmonary congestion: Post hoc subset analysis of the multicenter diltiazem postinfarction trial and the second Danish Verapamil Infarction Trial
studies. Am J Cardiol 86:275279, 2000
Jollis JG, Simpson RJ, Chowdhury MK, et al: Calcium channel
blockers and mortality in elderly patients with myocardial infarction. Arch Intern Med 159:23412348, 1999
Reicher-Reiss H, Behar S, Boyko V, et al: Long-term mortality
follow-up of hospital survivors of a myocardial infarction randomized to nifedipine in the SPRINT study. Secondary Prevention
Reinfarction Israeli Nifedipine Trial. Cardiovasc Drugs Ther 12:
171176, 1998
Moss AJ, Oakes D, Rubison M, et al: Effects of diltiazem on longterm outcome after acute myocardial infarction in patients with
and without a history of systemic hypertension. The Multicenter
Diltiazem Postinfarction Trial Research Group. Am J Cardiol 68:
429433, 1991
Agodoa LY, Appel L, Bakris GL, et al: Effect of ramipril vs

20.

21.
22.
23.
24.
25.

26.
27.
28.

29.
30.
31.
32.
33.
34.

amlodipine on renal outcomes in hypertensive nephrosclerosis: A

randomized controlled trial. JAMA 285:27192728, 2001
Goldbourt U, Behar S, Reicher-Reiss H, et al: Early administration of nifedipine in suspected acute myocardial infarction. The
Secondary Prevention Reinfarction Israel Nifedipine Trial 2 Study.
Arch Intern Med 153:345353, 1993
Ha SK, Park HS, Kim SJ, et al: Prevalence and patterns of left
ventricular hypertrophy in patients with predialysis chronic renal
failure. J Korean Med Sci 13:488494, 1998
Harnett JD, Parfrey PS, Griffiths SM, et al: Left ventricular
hypertrophy in end-stage renal disease. Nephron 48:107115, 1988
Ni Z, Wang XQ, Vaziri ND: Nitric oxide metabolism in erythropoietin-induced hypertension: Effect of calcium channel blockade.
Hypertens 32:724729, 1998
Vaziri ND, Ni Z, Wang XQ, et al: Downregulation of nitric oxide
synthase in chronic renal insufficiency: Role of excess PTH. Am
J Physiol 274:F642F649, 1998
De Rosa ML, Giordano A, Della Guardia D, et al: Reversal of
left ventricular hypertrophy following once daily administration
of felodipine for two years to elderly subjects with isolated systolic
hypertension. Cardiology 92:3944, 1999
Schmieder RE, Martus P, Klingbeil A: Reversal of left ventricular hypertrophy in essential hypertension. A meta-analysis of randomized double-blind studies. JAMA 275:15071513, 1996
Cice G, Tagliamonte E, Ferrara L, et al: Efficacy of diltiazem
in uremic hemodialyzed patients with isolated diastolic dysfunction
and dialysis hypotensive crisis. Cardiologia 44:289294, 1999
Thanakitcharu P, Fadda GZ, Hajjar SM, et al: Verapamil
prevents the metabolic and functional derangements in pancreatic islets of chronic renal failure rats. Endocrinology 129:1749
1754, 1991
Qing DP, Ding H, Vadgama J, et al: Elevated myocardial cytosolic
calcium impairs insulin-like growth factor-1-stimulated protein
synthesis in chronic renal failure. J Am Soc Nephrol 10:8492, 1999
Raine AE, Bedford L, Simpson AW, et al: Hyperparathyroidism,
platelet intracellular free calcium and hypertension in chronic renal
failure. Kidney Int 43:700705, 1993
Smogorzewski M, Tian J, Massry SG: Down-regulation of PTHPTHrP receptor of heart in CRF: Role of [Ca2]i. Kidney Int 47:
11821186, 1995
Picotto G, Massheimer V, Boland R: Parathyroid hormone stimulates calcium influx and the cAMP messenger system in rat enterocytes. Am J Physiol 273:C1349C1353, 1997
Skouteris GG: Parathyroid hormone co-stimulation of hepatocyte
proliferation is sensitive to protein kinase A and calcium channel
inhibitors. Biosci Rep 16:343350, 1996
Perneger TV, Klag MJ, Whelton PK: Cause of death in patients
with end-stage renal disease: death certificates vs. registry reports.
Am J Public Health 83:17351738, 1993