Original Article
Dialysate Sodium Prescription and Blood Pressure in
Hemodialysis Patients
Manfred Hecking,1 Angelo Karaboyas,2 Hugh Rayner,3 Rajiv Saran,4,5 Ananda Sen,6,7 Masaaki Inaba,8
Jürgen Bommer,9 Walter H. Hörl,1 Ronald L. Pisoni,2 Bruce M. Robinson,2 Gere Sunder-Plassmann,1
and Friedrich K. Port2
background
Diffusive sodium removal has been recommended to control hyperten-
sion in hemodialysis patients. Recent evidence on hospitalizations and
mortality, however, challenged the benefit of lower dialysate sodium
prescriptions and ignited a debate in the dialysis community. We there-
fore studied the relationship between dialysate sodium and blood pres-
sure over the longer term.
methods
We used multiply adjusted linear mixed models to estimate the associa-
tion between dialysate sodium and predialysis systolic blood pressure
(SBP) as well as change in SBP (delta SBP; postdialysis minus predialy-
sis) in 23,962 patients from the international Dialysis Outcomes and
Practice Patterns Study.
results
We found that 43% of hemodialysis facilities had variable (individual-
ized) dialysate sodium prescriptions (125–155 mEq/L), whereas 57%
had uniform dialysate sodium prescriptions (135–145 mEq/L) for
≥90% patients. Between-group comparisons of these 2 facility types
In the general population, the pathogenesis of hypertension
has been linked to higher sodium intake,1 with disagree-
ment on the association with cardiovascular events,2,3 as
well as the possible consequences regarding salt intake.4–6
In the hemodialysis population, even blood pressure man-
agement remains highly controversial,7–9 possibly because
epidemiological studies have failed to incriminate hyper-
tension as a cardiovascular risk factor10 and because defini-
tive clinical trials are missing.11 By conventional wisdom,
dietary sodium restriction, sodium removal by ultrafiltra-
tion, and diffusive sodium transport on dialysis have been
Correspondence: Friedrich K. Port (friedrich.port@arborresearch.org).
Initially submitted November 21, 2013; date of first revision December
13, 2013; accepted for publication February 5, 2014; online publication
March 20, 2014.
1160  American Journal of Hypertension  27(9)  September 2014
Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246
by guest
on 09 February 2018
suggested that dialysate sodium, when variably prescribed, might have
been used to modify predialysis SBP (Pinteraction = 0.01) and perhaps delta
SBP levels (Pinteraction = 0.08). Within facilities not prone to indication bias,
because dialysate sodium was not variable, higher uniform dialysate
sodium (per 2 mEq/L) was associated with slightly higher SBP (+0.9 mm
Hg, 95% confidence interval (CI) = 0.1–1.6 among all patients; +1.7 mm
Hg, 95% CI = 0.1–3.2 among patients not treated with blood pressure
medication) and no increase in delta SBP.
conclusions
Patients assigned to hemodialysis facilities with uniformly higher
dialysate sodium do not have markedly higher predialysis SBP, pro-
viding rather limited support for lowering dialysate sodium to control
hypertension, particularly in view of hospitalization and mortality risks
associated with lower dialysate sodium.
Keywords: blood pressure; dialysate sodium; hemodialysis; ­hypertension;
intradialytic hypotension; systolic blood pressure.
doi:10.1093/ajh/hpu040
considered key elements in control of hypertension in dialy-
sis patients.12–14
Sodium removal occurs during dialysis primarily by
convection through isotonic ultrafiltration. With respect
to diffusive sodium transport, intradialytic sodium load-
ing through high dialysate sodium (DNa) prescriptions has
been considered “evil”15 and recommended to be avoided,
per previous practice guidelines.16 A  strategy of aligning
DNa with serum sodium (SNa) concentration has been
advocated to reduce blood pressure, interdialytic weight
gain (IDWG), and intradialytic symptoms.17 Similarly,
1Department of Internal Medicine III–Nephrology, Medical University of
Vienna, Vienna, Austria; 2Arbor Research Collaborative for Health, Ann
Arbor, MI; 3Birmingham Heartlands Hospital, Birmingham; 4Department
of Internal Medicine–Nephrology, University of Michigan, Ann Arbor, MI;
5Kidney Epidemiology and Cost Center, University of Michigan, Ann Arbor,
MI; 6Department of Family Medicine, University of Michigan, Ann Arbor,
MI; 7Department of Biostatistics, University of Michigan, Ann Arbor, MI;
8Department of Metabolism, Endocrinology, and Molecular Medicine, Osaka
City University Graduate School of Medicine, Osaka, Japan; 9Department of
Nephrology, University of Heidelberg , Heidelberg, Germany.
© American Journal of Hypertension, Ltd 2014. All rights reserved.
For Permissions, please email: journals.permissions@oup.com

DNa matching to SNa has been promoted18 as a neglected
sodium restriction approach.19 Finally, the chief medical
officers of 14 major US dialysis providers have recently
recommended a reduction of DNa to the 134–138 mEq/L
range;20 this recommendation was subsequently coun-
tered.21 Importantly, limiting IDWG by restricting sodium
intake from diet or DNa does not guarantee achievement
of an optimal target weight (dry weight) at the end of
dialysis,10 which is central to good blood pressure control.
Moreover, in the early years of dialysis, when DNa pre-
scriptions were typically approximately 132 mEq/L, dialy-
sis disequilibrium syndrome was commonly observed;
it has become quite rare since the DNa was raised in the
1980s to approximately 140 mEq/L.21
Recent data from the international Dialysis Outcomes
and Practice Patterns Study (DOPPS) do not seem to jus-
tify the proposed recommendation to lower DNa. In our
previous DOPPS analyses,22 we confirmed a strong inverse
association between SNa and mortality23 and identified
a modest increase in IDWG with higher DNa prescrip-
tions but also a lower, not higher, hospitalization risk.24
We observed that approximately one-half of hemodialysis
facilities already varied (individualized) the DNa prescrip-
tion, albeit not usually to match SNa,24 whereas the other
half prescribed a uniform DNa for ≥90% of their patients.
In DOPPS facilities using variable DNa, the use of higher
DNa prescriptions was associated with patient comor-
bidities and consequently higher mortality rates despite
adjustment.24 In patients from facilities using uniform
DNa for ≥90% of their patients, uniformly higher DNa pre-
scriptions were associated with better survival, compared
with uniformly lower DNa,24 depending on the level of
adjustment.25
None of our previous analyses has focused on blood
pressure. Here, we estimated the associations between DNa
prescription and predialysis systolic blood pressure (SBP),
as well as change in SBP from pre- to post-dialysis (delta
SBP). While patients in individualized DNa facilities are
subject to confounding by indication, patients in uniform
DNa facilities are assigned quasi-randomly to the facility’s
uniform DNa without such confounding.24 Therefore, we
hypothesized that associations between DNa prescription
practice, SBP, and outcomes would differ between dialysis
facilities that individualized DNa prescription and facilities
that used a uniform DNa prescription.
METHODS
Data source
The DOPPS is a prospective cohort study that has enrolled
more than 50,000 representative patients in 12 countries.
More than 300 facilities were included in each of 4 com-
pleted study phases, DOPPS 1–4. Collected data include
patient demographics, dialysis prescription, medications,
laboratory measurements, and deaths, along with their
associated causes. Data are retrieved at baseline and every
4 months during follow-up.26,27 Detailed information on the
DOPPS study design and sampling scheme has been pub-
lished elsewhere.26,28
American Journal of Hypertension  27(9)  September 2014  1161
Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246
by guest
on 09 February 2018
Dialysate Sodium and Blood Pressure
Creation of the study cohort
Pre- and postdialysis SBP measurements, along with pre-
dialysis SNa and DNa prescription, were available for 67%
of eligible patients at their entry into the DOPPS (baseline
data). The percentage was 34% in DOPPS 1 (SNa not col-
lected during the initial facility round in the United States
and Japan), 74% in DOPPS 2, 82% in DOPPS 3, and 88% in
DOPPS 4. All patients were included in our analysis unless
they were on sodium modeling/profiling (n = 7,450; 14%),
had either pre- or postdialysis SBP measurements missing
(n  =  3,630; 7%), or had the following variables of interest
outside of the following boundaries: predialysis SNa outside
126–150 mEq/L (<1%), DNa prescription outside 125–155
mEq/L (<1%). Data from large dialysis organizations in
the United States in phase 4 were not included because of
incomplete data on comorbid conditions. To ensure that all
baseline data represented the steady state of the hemodialy-
sis population, patients who had started dialysis <30  days
before baseline were excluded.
Variables of interest
The DNa prescription was recorded in the DOPPS
patient-level data collection instruments at baseline and at
4-month follow-up intervals. A  hemodialysis facility was
considered to have individualized (varied) their DNa pre-
scription if >10% of that facility’s patients had a DNa that
was not equal to the DNa of the majority of its patients. All
other hemodialysis facilities were considered uniform DNa
facilities, but patients with a DNa prescription unequal to
the DNa of ≥90% of that facility’s patients were excluded
(<1%). Of patients with at least 2 reported follow-up val-
ues for DNa, 86% of patients overall and 97% of patients in
uniform DNa facilities had a mean follow-up DNa within 1
mEq/L of baseline DNa. Regarding predialysis SNa measure-
ments, recommendations from the International Federation
of Clinical Chemistry Expert Panel on conversion of sodium
levels should ensure that SNa values are very similar, whether
obtained by direct potentiometry or indirect potentiometry,
which is typically used in centralized laboratories.29 Baseline
SBP measurements in DOPPS 1, 2, and 4 were averages of
the 3 most recent readings before study entry. DOPPS 3
recorded only a single “most recent” measurement at base-
line. Overall, 65% of eligible patients had 3 measurements
available. Baseline delta SBP was defined as postdialysis SBP
minus predialysis SBP at baseline. Because directly meas-
ured data on IDWG were not always available, we estimated
IDWG as the measurements of baseline intradialytic weight
loss (expressed as a percentage of postdialysis weight and
normalized to mid-week), consistent with previous publica-
tions by us and others.30–32 Intradialytic weight loss corre-
lates strongly with the preceding IDWG (r = 0.84–0.87 based
on 7,878 patients in DOPPS 4).
Statistical analysis
Descriptive statistics were used to describe the prevalent
cross-section at baseline. Linear mixed models accounting
for clustering with a random facility intercept were used
 Hecking et al.
to regress predialysis SBP on DNa and to regress delta SBP
on DNa and predialysis SBP. Patients with DNa equal to
140 mEq/L were chosen as the reference group because they
were the largest category and thus best represented a typical
patient. Analyses were conducted separately among individ-
ualized DNa facilities and uniform DNa facilities.
Evaluation of factors associated with DNa prescription
was done separately for nonmodifiable and modifiable
patient characteristics because analyses of modifiable vari-
ables may be biased by DNa individualization. First, we used
adjusted linear mixed models among individualized DNa
facilities with DNa as the outcome and demographics and
comorbidities as covariates to assess which nonmodifiable
patient characteristics were associated with DNa prescrip-
tion. Next, we determined whether any modifiable variables
(baseline predialysis SBP, delta SBP, IDWG) were associ-
ated with the prescription practice of individualized DNa by
actually treating these variables as outcomes and DNa as the
predictor of interest. We tested whether the effect of DNa
differed between individualized and uniform DNa facilities
and thus was likely being prescribed on the basis of the vari-
able by evaluating the interaction between DNa and an indi-
cator for facility DNa individualization.
Sensitivity analyses were conducted (i) excluding Japan,
(ii) among patients not on antihypertensive medication, (iii)
among patients from hemodialysis facilities where 100%
of patients received a uniform DNa, (iv) among patients
with no residual kidney function (urine volume <200 ml/
day), and (v) by SNa strata. Adjustments for models with
predialysis SBP and delta SBP as an outcome variable are
described in the figure captions. Although adjustments for
SNa were appropriate in our analyses stratified by DNa,
additional adjustments for the dialysate-to-serum sodium
gradient (GNa) would not have been appropriate, because
GNa depends on both DNa and SNa (GNa = DNa – SNa).
An extensive examination of the role GNa is described
elsewhere.24
Missing covariate values were multiply imputed using the
chained equation method33 by IVEWARE.34 Results from
5 imputed datasets were combined for the final analysis
using Rubin’s formula,35 which was implemented in PROC
MIANALYZE in SAS. The proportion of missing data was
<10% for all imputed covariates with the exception of fer-
ritin (20%) and IDWG (10%). For calculations, we used SAS
9.2 (SAS Institute, Cary, NC).
RESULTS
Twenty-three thousand nine hundred sixty-two hemodi-
alysis patients from DOPPS phases 1–4 were studied in this
analysis. Patient characteristics of the study population cat-
egorized by 3 strata of DNa are listed in Table 1. The DNa
prescription varied markedly by country and ranged 125–
155 mEq/L in 401 individualized DNa facilities (43% of all
facilities) and ranged 135–145 mEq/L in 528 uniform DNa
facilities (57% of all facilities) (Figure  1). Among all indi-
vidualized DNa facilities, the highest percentage of patients
with 1 specific DNa was 75%–90% in 140 (35%) facilities,
50%–75% in 178 (44%) facilities, and <50% in 83 (21%)
facilities.
1162  American Journal of Hypertension  27(9)  September 2014
Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246
by guest
on 09 February 2018
The overall mean predialysis SBP was lower in patients
from individualized DNa facilities (140.7 mm Hg) com-
pared with uniform DNa facilities (146.2 mm Hg). However,
3.7 mm Hg of this 5.5 mm Hg crude difference between the
2 groups was attributable to country differences, which were
adjusted for in all models. Other crude differences in patient
characteristics (Table 1) were likely confounded by country
and therefore not interpreted further.
Association between DNa prescription and predialysis SBP
Predialysis SBP was lower in patients who were pre-
scribed higher DNa concentrations (Figure 2). Continuously
described, predialysis SBP was 1.7 mm Hg lower per 2 mEq/L
higher DNa (95% confidence interval (CI) = −2.1 to −1.2).
This association is in contrast with the results of small inter-
ventional trials on DNa reduction.36 We therefore explored
the role of treatment practices in this association. In patients
from individualized DNa facilities susceptible to confound-
ing by indication, predialysis SBP was 2.0 mm Hg lower per
2 mEq/L higher individualized DNa (95% CI = −2.5 to −1.5)
(Figure 2). By contrast, in patients from dialysis facilities that
prescribed a uniform DNa in ≥90% of their patients, predial-
ysis SBP was 0.9 mm Hg higher per 2 mEq/L higher uniform
DNa (95% CI = 0.1–1.6) (Figure 2).
Association between DNa prescription and pre- to
postdialysis delta SBP
In individualized DNa facilities, delta SBP was 1.1 mm
Hg more negative per 2 mEq/L higher DNa (95% CI = −1.4
to −0.7) (Figure 3). When restricting to uniform DNa facili-
ties, higher uniform DNa prescription was not strongly
associated with delta SBP (−0.4 mm Hg per 2 mEq/L higher
uniform DNa; 95% CI = −1.0 to 0.2). Because the delta SBP
drop (from pre- to postdialysis) is expected to be larger
in patients with higher predialysis SBP (Supplementary
Figure S1), we adjusted the aforementioned analyses for
predialysis SBP.
Determinants of individualized DNa prescription
Prescription of individualized DNa was associated with
only a few of the nonmodifiable demographics and comor-
bidities tested, led by age (<0.1 mEq/L higher per 10 years
older; data not shown). Tests of modifiable patient charac-
teristics are shown in Table 2; the results for predialysis SBP
and delta SBP were congruent with Figure 2 and Figure 3,
respectively. Higher DNa was significantly associated with
lower predialysis SBP in individualized DNa facilities and
with higher predialysis SBP in uniform DNa facilities. The
significant interaction (P = 0.01) indicated that DNa might
have been individually prescribed according to a patient’s
predialysis SBP, specifically higher DNa prescription for
patients with lower SBP and vice versa. For delta SBP, the
interaction was of borderline significance (P  =  0.08), sug-
gesting that patients prone to larger drops in SBP might
have been prescribed an individualized higher DNa. For
IDWG, the estimated effect of DNa was very similar among
by guest
Table 1.  Patient characteristics by categories of dialysate sodium prescription

on 09 February 2018
All patients Patients in individualized DNa facilities Patients in uniform DNa facilities

Baseline characteristic DNa < 140 DNa = 140 DNa > 140 DNa < 140 DNa = 140 DNa > 140 DNa < 140 DNa = 140 DNa > 140

No. of facilities — — — — — — 110 367 51

No. of patients 6,472 13,834 3,656 3,576 4,260 2,256 2,827 9,387 1,344
Men 59 60 58 59 60 57 58 60 60
Urine output >200 ml/day 45 34 29 45 41 29 45 31 30
Diabetes mellitus 39 38 32 39 37 30 38 39 35
Black race 7 6 4 5 6 5 9 6 4
Comorbid conditions
 Hypertension 85 81 77 85 84 77 83 79 78
  Coronary artery disease 43 41 39 45 43 41 41 41 35
  Cardiovascular disease 34 35 37 35 38 41 31 34 31
  Congestive heart failure 30 30 27 30 29 29 28 30 23
  Peripheral vascular disease 28 26 26 30 32 29 27 24 22
  Cerebrovascular disease 17 17 16 17 17 18 17 16 14
  Psychiatric disorder 18 15 16 20 18 18 16 14 11

Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246

  Cancer, nonskin 13 13 12 14 15 13 12 12 11
  Lung disease 13 11 10 14 14 13 11 10 6
  Neurologic disease 11 11 11 12 12 12 11 10 10
  Recurrent cellulitis 10 8 7 11 9 8 9 7 5
  Gastrointestinal bleeding 6 5 5 6 6 5 6 5 5
 HIV/AIDS 1 1 1 0 1 1 1 1 0
Vascular access
  Arteriovenous fistula 63 67 76 64 63 73 62 69 81
  Dialysis catheter 23 18 12 23 23 14 22 16 7
 Graft 9 9 9 8 9 9 11 9 8
  Unknown access 5 6 4 4 5 4 5 6 4
Age, y 62.7 ± 15.1 62.9 ± 14.4 63.7 ± 14.1 62.7 ± 15.1 62.9 ± 15.1 64.7 ± 14.2 62.9 ± 15.0 63.0 ± 14.1 62.0 ± 13.7
Dialysis vintage, years 4.0 ± 5.1 4.9 ± 6.0 5.9 ± 6.6 4.0 ± 5.1 4.4 ± 5.6 5.7 ± 6.6 4.1 ± 5.1 5.1 ± 6.2 6.4 ± 6.7
Body mass index, kg/m2 25.8 ± 5.6 24.4 ± 5.7 23.7 ± 5.3 25.8 ± 5.7 25.5 ± 5.7 24.5 ± 5.4 25.7 ± 5.5 23.8 ± 5.6 22.3 ± 4.8

(Continued)

American Journal of Hypertension  27(9)  September 2014  1163

Dialysate Sodium and Blood Pressure
by guest
Table 1.  Continued

All patients Patients in individualized DNa facilities Patients in uniform DNa facilities

on 09 February 2018
Baseline characteristic DNa < 140 DNa = 140 DNa > 140 DNa < 140 DNa = 140 DNa > 140 DNa < 140 DNa = 140 DNa > 140
Hecking et al.

Interdialytic weight gain, % body weight 2.7 ± 1.8 3.3 ± 1.8 3.7 ± 1.8 2.6 ± 1.8 3.0 ± 1.8 3.7 ± 1.8 2.7 ± 1.8 3.4 ± 1.8 3.8 ± 1.8
Dialysis treatment time, minutes 241 ± 39 235 ± 34 235 ± 31 244 ± 37 237 ± 37 236 ± 33 238 ± 40 235 ± 32 233 ± 28
SBP
  Predialysis SBP, mm Hg 142.6 ± 22.8 144.5 ± 22.7 143.8 ± 24.2 141.4 ± 22.4 140.6 ± 22.1 139.6 ± 24.4 144.2 ± 23.2 146.1 ± 22.8 150.7 ± 22.1
  Postdialysis SBP, mm Hg 134.9 ± 23.2 135.0 ± 23.1 132.2 ± 23.3 134.2 ± 23.2 132.4 ± 22.6 128.6 ± 23.5 135.7 ± 23.2 136.1 ± 23.2 137.8 ± 21.5
  Delta SBP, mm Hg −7.7 ± 19.7 −9.5 ± 20.6 −11.6 ± 19.5 −7.2 ± 19.6 −8.2 ± 19.4 −11.0 ± 19.4 −8.5 ± 19.9 −10.1 ± 21.1 −12.9 ± 19.5
Laboratory measurements
  Serum albumin, g/dl 3.7 ± 0.5 3.7 ± 0.5 3.7 ± 0.5 3.7 ± 0.5 3.7 ± 0.5 3.7 ± 0.5 3.7 ± 0.5 3.7 ± 0.5 3.7 ± 0.5
  Hemoglobin, g/dl 11.3 ± 1.6 11.1 ± 1.6 11.0 ± 1.5 11.4 ± 1.5 11.3 ± 1.5 11.2 ± 1.5 11.2 ± 1.6 11.0 ± 1.6 10.7 ± 1.4
  Serum creatinine, mg/dl 8.3 ± 2.9 9.1 ± 3.1 9.3 ± 3.0 8.3 ± 2.9 8.5 ± 2.9 9.0 ± 2.9 8.3 ± 2.8 9.3 ± 3.1 9.9 ± 3.1
  Serum ferritin, ng/ml 467 ± 425 405 ± 413 416 ± 502 472 ± 459 429 ± 405 448 ± 488 461 ± 378 394 ± 417 362 ± 532
  WBC count, 1,000/mm3 7.2 ± 2.4 6.9 ± 2.4 6.7 ± 2.3 7.2 ± 2.3 7.1 ± 2.4 6.9 ± 2.3 7.2 ± 2.4 6.7 ± 2.3 6.4 ± 2.2
  Predialysis SNa, mEq/L 138.0 ± 3.5 138.4 ± 3.4 138.4 ± 3.4 138.0 ± 3.5 138.2 ± 3.4 138.3 ± 3.5 138.1 ± 3.6 138.5 ± 3.3 138.6 ± 3.2
Dialysate Na prescription, mEq/L 137.8 ± 1.0 140.0 ± 0.0 142.8 ± 1.7 137.7 ± 1.2 140.0 ± 0.0 143.0 ± 2.0 138.0 ± 0.8 140.0 ± 0.0 142.6 ± 0.8

Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246

1164  American Journal of Hypertension  27(9)  September 2014
Dialysate-to-serum Na gradient, mEq/L −0.2 ± 3.6 1.6 ± 3.4 4.5 ± 3.7 −0.3 ± 3.6 1.8 ± 3.4 4.7 ± 3.9 −0.1 ± 3.7 1.5 ± 3.3 4.0 ± 3.2

Data are mean ± SD or percentage unless otherwise noted. Columns do not sum because only those facilities with sufficient data (n ≥ 5 patients) were included in facility analyses. Delta
systolic blood pressure (intradialytic change in systolic blood pressure) denotes postdialysis minus predialysis systolic blood pressure.
Abbreviations: DNa, dialysis sodium prescription; Na; sodium; SBP, systolic blood pressure; SNa, serum sodium; WBC, white blood cell.
 Dialysate Sodium and Blood Pressure

Figure  1.  Distribution of facility dialysate sodium prescription, as practiced in uniform dialysate sodium (DNa) facilities by Dialysis Outcomes and
Practice Patterns Study (DOPPS) country. The number of total facilities (n= 528) is calculated by adding the number of participating facilities per study
phase; these facilities are represented by 302 unique facilities across DOPPS phases. Abbreviations: ANZ, Australia/New Zealand; BE, Belgium; CA, Canada;
FR, France; GE, Germany; IT, Italy; JP, Japan; SP, Spain; SW, Sweden; UK, United Kingdom; US, United States.

individualized and uniform DNa facilities (Pinteraction = 0.90),

suggesting that DNa was not tailored in response to IDWG.

Antihypertensive medication

Seventy percent of patients were receiving at least 1 antihy-

pertensive medication at baseline. Restricting to patients in
uniform DNa facilities not receiving antihypertensives, predi-
alysis SBP was 1.7 mm Hg higher per 2 mEq/L higher uniform
DNa (95% CI = 0.1–3.2). This association of DNa with predi-
alysis SBP in uniform DNa facilities was stronger than among
patients treated with antihypertensives (Pinteraction  =  0.006)
(Figure  4). Also, among patients in uniform DNa facilities
not receiving antihypertensives, delta SBP was slightly more
negative for patients with higher DNa (−0.8 mm Hg per 2
mEq/L higher uniform DNa; 95% CI = −1.9 to +0.3).
Other sensitivity analyses
A sensitivity analysis testing the association between DNa
prescription and predialysis SBP in individualized vs. uni-
form DNa facilities, but excluding Japan, given the unique-
ness of the Japanese hemodialysis population, yielded results
that were consistent with Figure 2 and Table 2. In a sensi-
tivity analysis defining a uniform facility as 100% (rather
than ≥90%) of patients prescribed the same DNa (41% of
facilities), the association between DNa and predialysis SBP
was weak (0.2 mm Hg higher per 2 mEq/L higher uniform
DNa). A sensitivity analysis among the 64% of patients with
no residual kidney function (urine volume <200 ml/day)
showed similar results to the main analysis. When patients
were stratified by their predialysis SNa, the association
between DNa and SBP within both individualized DNa and
uniform DNa facilities did not vary meaningfully by cate-
gory of SNa (Supplementary Figure S2).
Figure 2.  Predialysis systolic blood pressure (SBP) and dialysate sodium
(DNa) prescription in hemodialysis patients. Three separate linear regres-
sion models using reference point of 140 mEq/L. All models adjusted
for Dialysis Outcomes and Practice Patterns Study (DOPPS) phase and
country, age, sex, black race, vintage, body mass index, diabetes (as
comorbidity or cause of end-stage renal disease), 12 other comorbidities
(not including hypertension), residual renal function (i.e., urine output
>200 ml/day), interdialytic weight gain, vascular access, serum albumin,
serum creatinine, hemoglobin, ferritin, white blood cell count, serum
sodium, and 3 facility practice indicators (% catheter use, % spKt/V <1.2,
and % serum phosphorus ≥5.5 mg/dl). P values represent the test for lin-
ear trend (i.e., models expressing DNa as a continuous variable). The lin-
ear effect of DNa was significantly different in individualized vs. uniform
facilities (Pinteraction = 0.01).
Discussion
The principal findings in this study are (i) the practice of
varying (individualizing) the DNa prescription appeared to be
related to the management of blood pressure, with lower indi-
vidualized DNa prescribed to patients with higher predialysis
SBP, with a suggestion of higher individualized DNa prescribed
to patients with larger drops in delta SBP (and vice versa);

American Journal of Hypertension  27(9)  September 2014  1165

Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246

by guest
on 09 February 2018
 Hecking et al.

Figure 3.  Delta systolic blood pressure (SBP; post- minus predialysis) and dialysate sodium (DNa) prescription in hemodialysis patients. Two separate
linear regression models using reference point of DNa = 140 mEq/L. All models adjusted for Dialysis Outcomes and Practice Patterns Study (DOPPS)
phase and country, age, sex, black race, vintage, body mass index, diabetes (as comorbidity or cause of end-stage renal disease), 12 other comorbidities
(not including hypertension), residual renal function (i.e., urine output >200 ml/day), interdialytic weight gain, vascular access, serum albumin, creati-
nine, hemoglobin, ferritin, white blood cell count, serum sodium, predialysis systolic blood pressure, and 3 facility practice indicators (% catheter use, %
spKt/V <1.2, and % serum phosphorus ≥5.5 mg/dl). P values represent the test for linear trend using DNa as a continuous variable. The linear effect of DNa
showed a trend to be different in individualized vs. uniform facilities (Pinteraction = 0.08).

Table 2.  Modifiers of the dialysate sodium (DNa) prescription in individualized DNa facilities using uniform DNa facilities as control, with
assessment of interactions

Model 1: Outcome per 2 mEq/L higher Model 2: Outcome per 2 mEq/L higher DNa in Overall, Model 3:
Model outcome DNa in individualized DNa facilities uniform DNa facilities Pinteractiona

Predialysis SBP, mm Hg −2.01 (−2.48 to −1.54) 0.85 (0.13 to 1.58) 0.01

Delta SBP, mm Hg −1.07 (−1.44 to −0.70) −0.38 (−0.97 to 0.21) 0.08
IDWG, % of body weight 0.10 (0.07 to 0.14) 0.14 (0.08 to 0.19) 0.93

Linear mixed models: Estimated difference (95% confidence interval) in the outcome variable is shown per 2 mEq/L higher dialysate sodium
(DNa). Nine separate models with DNa as the exposure and the outcome as listed above, adjusted for Dialysis Outcomes and Practice Patterns
Study (DOPPS) phase and country, age, sex, black race, vintage, body mass index, diabetes (as comorbidity or cause of end-stage renal
disease), 12 other comorbidities (not including hypertension), residual renal function (i.e., urine output >200 ml/day), interdialytic weight gain,
vascular access, serum albumin, creatinine, hemoglobin, ferritin, white blood cell count, serum sodium, predialysis systolic blood pressure, and
3 facility practice indicators (% catheter use, % spKt/V <1.2, and % serum phosphorus ≥5.5 mg/dl).
aInteraction of DNa and indicator for whether a facility DNa was individualized.

(ii) in facilities where the DNa prescription was not variably
prescribed (uniform), higher uniform DNa was associated with
a slightly higher predialysis SBP and no substantial difference
in delta SBP. Thus, the results from uniform DNa facilities show
only a small effect of DNa on predialysis SBP and therefore
question the practice of varying DNa based on this reading.
As in our previous study,24 we judge the analysis of
patients in uniform DNa facilities to have characteristics
of a quasi-randomized experiment in which patients with
a range of comorbidities are allocated to a dialysis facil-
ity principally because of its proximity to their residence
and then subjected to a single DNa chosen by the facility.
By contrast, in individualized DNa facilities, DNa is often
prescribed by indication based on predialysis SBP, which
may correlate with comorbidities and additional patient
characteristics; therefore, associations are likely biased by
indication, for which it may not be possible to adjust fully.
In intervention studies of DNa reduction, blood pres-
sure decreased significantly and by a greater magnitude
than observed in this study, at least when considering those
DOPPS patients receiving antihypertensive medication (the
large majority).36 These intervention trials were relatively
short-term (up to 16 weeks), and at least 1 of the trials simul-
taneously reduced dietary sodium intake. Furthermore, the
DNa reductions were substantial (5 mEq/L in most studies).
Recently, Zhou et  al. showed blood pressure and IDWG
declined with lower DNa in patients at the same volume sta-
tus.37 Suckling et al. showed a decrease in blood pressure when
using DNa 135 rather than 145 mEq/L but did not report on
symptoms,38 whereas Kim et al. showed a decrease in blood
pressure and IDWG with lower DNa, as well as a doubling in
the crude frequency of cramps and headaches, which was not
statistically significant in their study of 32 patients.39
Shah et  al. showed statistically more intradialytic hypo-
tensive episodes per week with DNa ≤136 mEq/L in compar-
ison with DNa 137–139 mEq/L and DNa ≥140 mEq/L, and
with DNa ≥140 mEq/L in comparison with DNa 137–139
mEq/L and DNa ≤136 mEq/L.40 They commented on higher

1166  American Journal of Hypertension  27(9)  September 2014

Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246

by guest
on 09 February 2018
 Dialysate Sodium and Blood Pressure

Figure 4.  Dialysate sodium (DNa) and predialysis systolic blood pressure (SBP) by blood pressure (BP) medication. Four separate linear regression mod-
els using reference point of DNa = 140 mEq/L, adjusted for Dialysis Outcomes and Practice Patterns Study (DOPPS) phase and country, age, sex, black race,
vintage, body mass index, diabetes (as comorbidity or cause of end-stage renal disease), 12 other comorbidities (not including hypertension), residual
renal function (i.e., urine output >200 ml/day), interdialytic weight gain, vascular access, serum albumin, creatinine, hemoglobin, ferritin, white blood cell
count, serum sodium, and 3 facility practice indicators (% catheter use, % spKt/V <1.2, and % serum phosphorus ≥5.5 mg/dl).

SBP and increased prescription of antihypertensive medi-
cations in those patients dialyzing against the lower rather
than higher DNa concentrations, supporting our suggestion
that bias by indication findings associated with individual-
ized DNa prescription.
In a recent large intervention study, physicians were given
the option to choose between a facility-wide DNa prescrip-
tion of 137 mEq/L vs. retaining 140 mEq/L. In the lower DNa
group, only a very small but statistically significant reduction
in SBP was observed (by 0.65 mm Hg),41 which agrees with
the results of this study.
Numerous differences in study design may explain some
discrepancies with the reports here above. For example, the
present study analyzed cross-sectional data and excluded
patients who had been on dialysis for <30  days to allow
reaching a relative steady state of salt and water balance at
baseline. Titration of blood pressure medication doses to
desired SBP targets may have attenuated associations so that
our subanalyses restricted to patients not on antihyperten-
sives are cleaner and more relevant (Figure 4).
Nursing interventions against intradialytic hypotension fre-
quently increase sodium transfer to the patient, thereby possi-
bly increasing SBP. Our finding that higher individualized DNa
prescriptions were associated with a larger drop to postdialysis
SBP (delta SBP) in patients from individualized DNa facilities
suggests that dialysis physicians may make use of the DNa to try
to reduce intradialytic hypotension. In uniform DNa facilities,
we did not identify a significant inverse association between
higher DNa prescription and delta SBP, suggesting that delta
SBP was not less negative with higher DNa. Unfortunately, we
did not have data on blood pressure readings during dialysis
or treatment for intradialytic hypotension. Increases in the
DNa concentration have been found to be effective in reducing
dialysis disequilibrium syndrome, which was common when
a DNa prescription of 132 mEq/L was standard, but this syn-
drome did not typically include hypotension.42
In this observational study, we use the term “individual-
ized DNa” even though we cannot prove that the DNa had
been intentionally individualized. The identified associations
can only suggest that DNa was prescribed in response to the
patient’s level of blood pressure control. It is possible that higher
DNa may have been used in some facilities to treat sicker or
unstable patients,24 many of whom have lower predialysis SBP
readings or a greater delta SBP. Clinical practice guidelines
included the DNa prescription practice in recommendations
for blood pressure management in hemodialysis patients by
advising to avoid a positive sodium balance induced by hyper-
tonic dialysate.16 The term “individualization” for dialysate
modification and DNa prescription has been used in numer-
ous previous original publications43 and recommendations,12,18
albeit mostly in the context with lowering the DNa.36
Our study shows an overall effect of DNa on blood pres-
sure, but it seems not as strong as might be expected. It
therefore appears to be of limited value to alter the DNa pre-
scription to influence either high or low predialysis blood
pressure and possibly intradialytic hypotension. We have
previously uncovered that the magnitude of IDWG associ-
ated with the DNa prescription appears to be much smaller
than expected from smaller, previous studies.24 Moreover,
analyses of morbidity and mortality showed that lower, not
higher, DNa prescriptions are associated with adverse out-
comes in uniform DNa facilities.24 Finally, patients who
received lower DNa prescriptions in the DOPPS reported
that they required longer time to recover from the dialysis
session and that they were more frequently bothered by mus-
cle cramps.44 In a recent sensitivity analysis of our data where
we excluded Japan and further added adjustments for other
facility practices, we saw that the mortality benefit of higher
DNa prescriptions in uniform DNa facilities remained, albeit
somewhat attenuated.25 If anything, however, the previous
DOPPS analyses, taken together, argue against the recent
suggestions that prescribing slightly higher DNa is harmful

American Journal of Hypertension  27(9)  September 2014  1167

Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246

by guest
on 09 February 2018
 Hecking et al.
because we reported that patients have better self-reported
outcomes and tend to have lower morbidity and mortality.
A strength of this study is that our large dataset allowed us
to use extensive case-mix adjustments, and our approach of
assigning patients to uniform DNa practices helped to over-
come the important bias by indication for different DNa pre-
scriptions. We demonstrated the value of this quasi-random
assignment of patients to a uniform DNa by comparisons
with results from dialysis facilities that varied the DNa.
Limitations include the fact that interventions during
dialysis were not recorded in the DOPPS and DNa was
not directly measured, as this is prescribed based on sup-
plier information and rarely confirmed in clinical practice.
Laboratory methods for SNa measurement may have varied
but are likely comparable regardless of the measurement
technique.29 Predialysis SBP levels are presented here as they
are used commonly in clinical practice, even though most
SBP data were averages from 3 readings during 1 week at
baseline (DOPPS 1, 2, and 4). Nevertheless, pre- and post-
dialysis SBP measurements may have remained imprecise
and inferior to ambulatory blood pressure, which has been
shown to enable a refined cardiovascular risk assessment.45,46
Any imprecision of our measurements (DNa, SNa, SBP,
delta SBP), however, is likely not biased in any direction and
would therefore tend to weaken our ability to find significant
results (e.g., have a type II error). Because we were neverthe-
less able to show clear associations, refined measurements
might not systematically have altered the direction of our
findings. Despite the described limitations, we therefore sug-
gest that our dataset provides information that is clinically
useful as it deals with the practice of prescribing DNa.
In conclusion, this study shows that individualized DNa
prescriptions appear to be used for blood pressure manage-
ment, with higher DNa prescribed for patients with lower
predialysis SBP and greater intradialytic drops in SBP.
Higher DNa in uniform DNa facilities is significantly asso-
ciated with a slightly higher predialysis SBP. These findings
provide very little clinically relevant support for the practice
of lowering DNa to decrease high SBP or for the practice of
using higher DNa to avoid a fall in SBP during hemodialysis.
Our previous studies on SNa,22 DNa and IDWG,24 morbid-
ity and mortality,24,25 and patient-reported outcomes44 sug-
gest that higher DNa prescriptions are potentially beneficial,
and, if anything, not harmful, independent of indication
bias. Until results from randomized controlled clinical tri-
als become available, the present focus on lowering the DNa
prescription does not appear to be justified.
SUPPLEMENTARY MATERIAL
Supplementary materials are available at American Journal
of Hypertension (http://ajh.oxfordjournals.org).
Acknowledgments
The DOPPS is administered by Arbor Research
Collaborative for Health and is supported by scientific
1168  American Journal of Hypertension  27(9)  September 2014
Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246
by guest
on 09 February 2018
research grants from Amgen (since 1996), Kyowa Hakko
Kirin (since 1999, in Japan), Sanofi Renal (since 2009),
AbbVie (since 2009), Baxter (since 2011), and Vifor Fresenius
Renal Pharma (since 2011), without restrictions on publi-
cations. For this analysis, Dr Hecking received a research
grant from the Else Kröner-Fresenius Foundation without
restrictions on publications and has been a visiting scientist
at Arbor Research Collaborative for Health. Heather Van
Doren, senior medical editor, provided editorial assistance
for this manuscript; Keith McCullough, senior research ana-
lyst, was consulted for additional statistical support (both
employees of Arbor Research Collaborative for Health).
Dr Walter Hörl is recenetly deceased; he was a significant
contributor to the article.
DISCLOSURE
The authors declared no conflict of interest.
References
1. Adrogue HJ, Madias NE. Sodium and potassium in the pathogenesis of
hypertension. N Engl J Med 2007; 356:1966–1978.
2. Bibbins-Domingo K, Chertow GM, Coxson PG, Moran A, Lightwood
JM, Pletcher MJ, Goldman L. Projected effect of dietary salt reductions
on future cardiovascular disease. N Engl J Med 2010; 362:590–599.
3. Stolarz-Skrzypek K, Kuznetsova T, Thijs L, Tikhonoff V, Seidlerova J,
Richart T, Jin Y, Olszanecka A, Malyutina S, Casiglia E, Filipovsky J,
Kawecka-Jaszcz K, Nikitin Y, Staessen JA, European Project on Genes
in Hypertension I. Fatal and nonfatal outcomes, incidence of hyperten-
sion, and blood pressure changes in relation to urinary sodium excre-
tion. JAMA 2011; 305:1777–1785.
4. Kotchen TA. The salt discourse in 2013. Am J Hypertens 2013; 26:1177.
5. Clapp JE, Curtis CJ, Kansagra SM, Farley TA. Getting the mes-
sage right: reducing sodium intake saves lives. Am J Hypertens 2013;
26:1181–1182.
6. McCarron DA. Physiology, not policy, drives sodium intake. Am J
Hypertens 2013; 26:1191–1193.
7. Mailloux LU, Haley WE. Hypertension in the ESRD patient: pathophys-
iology, therapy, outcomes, and future directions. Am J Kidney Dis 1998;
32:705–719.
8. Levin NW, Kotanko P, Eckardt KU, Kasiske BL, Chazot C, Cheung
AK, Redon J, Wheeler DC, Zoccali C, London GM. Blood pressure
in chronic kidney disease stage 5d-report from a kidney disease:
improving global outcomes controversies conference. Kidney Int 2010;
77:273–284.
9. Robinson BM, Tong L, Zhang J, Wolfe RA, Goodkin DA, Greenwood
RN, Kerr PG, Morgenstern H, Li Y, Pisoni RL, Saran R, Tentori F,
Akizawa T, Fukuhara S, Port FK. Blood pressure levels and mortal-
ity risk among hemodialysis patients in the Dialysis Outcomes and
Practice Patterns Study. Kidney Int 2012; 82:570–580.
10. Agarwal R, Weir MR. Dry-weight: a concept revisited in an effort to
avoid medication-directed approaches for blood pressure control in
hemodialysis patients. Clin J Am Soc Nephrol 2010; 5:1255–1260.
11. Horl WH. Hypertension in end-stage renal disease: different meas-
ures and their prognostic significance. Nephrol Dial Transplant 2010;
25:3161–3166.
12. Flanigan MJ. How should dialysis fluid be individualized for the chronic
hemodialysis patient? Sodium. Semin Dial 2008; 21:226–229.
13. Lindley EJ. Reducing sodium intake in hemodialysis patients. Semin
Dial 2009; 22:260–263.
14. Scribner BH, Buri R, Caner JE, Hegstrom R, Burnell JM. The treatment
of chronic uremia by means of intermittent hemodialysis: a preliminary
report. Trans Am Soc Artif Intern Organs 1960; 6:114–122.
15. Thijssen S, Raimann JG, Usvyat LA, Levin NW, Kotanko P. The evils of
intradialytic sodium loading. Contrib Nephrol 2011; 171:84–91.

16. Jindal K, Chan CT, Deziel C, Hirsch D, Soroka SD, Tonelli M, Culleton
BF, Canadian Society of Nephrology Committee for Clinical Practice
G. Hemodialysis clinical practice guidelines for the Canadian Society
of Nephrology. J Am Soc Nephrol 2006; 17:S1–S7.
17. Penne EL, Sergeyeva O. Sodium gradient: a tool to individualize
dialysate sodium prescription in chronic hemodialysis patients? Blood
Purif 2011; 31:86–91.
18. Lomonte C, Basile C. Do not forget to individualize dialysate sodium
prescription. Nephrol Dial Transplant 2011; 26:1126–1128.
19. Munoz Mendoza J, Sun S, Chertow GM, Moran J, Doss S, Schiller B.
Dialysate sodium and sodium gradient in maintenance hemodialysis: a
neglected sodium restriction approach? Nephrol Dial Transplant 2011;
26:1281–1287.
20. Parker T, 3rd. Creating an open dialogue on improving dialysis care.
Nephrol News Issues 2013; 27:14,16.
21. Port F, Hecking M, Karaboyas A, Pisoni R, Robinson B. Current evi-
dence argues against lowering the dialysate sodium. Nephrol News
Issues 2013 December; 27(13):18–21.
22. Hecking M, Karaboyas A, Saran R, Sen A, Horl WH, Pisoni RL,
Robinson BM, Sunder-Plassmann G, Port FK. Predialysis serum
sodium level, dialysate sodium, and mortality in maintenance hemo-
dialysis patients: the Dialysis Outcomes and Practice Patterns Study
(DOPPS). Am J Kidney Dis 2012; 59:238–248.
23. Waikar SS, Curhan GC, Brunelli SM. Mortality associated with low
serum sodium concentration in maintenance hemodialysis. Am J Med
2011; 124:77–84.
24. Hecking M, Karaboyas A, Saran R, Sen A, Inaba M, Rayner H, Horl
WH, Pisoni RL, Robinson BM, Sunder-Plassmann G, Port FK.
Dialysate sodium concentration and the association with interdialytic
weight gain, hospitalization, and mortality. Clin J Am Soc Nephrol 2012;
7:92–100.
25. Hecking M, Karaboyas A, Port FK. Dialysate sodium and mortality.
Letter to the Editor. Nephrol Dial Transplant, published online October
2012.http://ndt.oxfordjournals.org/content/27/4/1613/reply. Accessed
3 October 2014.
26. Pisoni RL, Gillespie BW, Dickinson DM, Chen K, Kutner MH, Wolfe
RA. The Dialysis Outcomes and Practice Patterns Study (DOPPS):
design, data elements, and methodology. Am J Kidney Dis 2004; 44:7–15.
27. Robinson BM, Port FK. Caring for dialysis patients: international
insights from the Dialysis Outcomes and Practice Patterns Study
(DOPPS). Identifying best practices and outcomes in the DOPPS.
Semin Dial 2010; 23:4–6.
28. Young EW, Goodkin DA, Mapes DL, Port FK, Keen ML, Chen K,
Maroni BL, Wolfe RA, Held PJ. The Dialysis Outcomes and Practice
Patterns Study (DOPPS): an international hemodialysis study. Kidney
Int 2000; 57:S74–S81.
29. Holbek CC. Understanding the different values in electrolyte measure-
ments. http://acutecaretesting.org.
30. Saran R, Bragg-Gresham JL, Rayner HC, Goodkin DA, Keen ML, Van
Dijk PC, Kurokawa K, Piera L, Saito A, Fukuhara S, Young EW, Held PJ,
Port FK. Nonadherence in hemodialysis: associations with mortality,
hospitalization, and practice patterns in the DOPPS. Kidney Int 2003;
64:254–262.
American Journal of Hypertension  27(9)  September 2014  1169
Downloaded from https://academic.oup.com/ajh/article-abstract/27/9/1160/2743246
by guest
on 09 February 2018
Dialysate Sodium and Blood Pressure
31. Leggat JE, Jr., Orzol SM, Hulbert-Shearon TE, Golper TA, Jones CA,
Held PJ, Port FK. Noncompliance in hemodialysis: predictors and sur-
vival analysis. Am J Kidney Dis 1998; 32:139–145.
32. Tang HL, Wong SH, Chu KH, Lee W, Cheuk A, Tang CM, Kong IL, Fung
KS, Tsang WK, Chan HW, Tong KL. Sodium ramping reduces hypotension
and symptoms during haemodialysis. Hong Kong Med J 2006; 12:10–14.
33. van Buuren S, Boshuizen HC, Knook DL. Multiple imputation of
missing blood pressure covariates in survival analysis. Stat Med 1999;
18:681–694.
34. Raghunathan TE, Solenberger PW, Van Hoewyk J: IVEware: Imputation
and variance estimation software. Survey Methodology Program,
Survey Research Center, Institute for Social Research, University of
Michigan: Ann Arbor, 2002.
35. Little RJA, Rubin DB. Statistical Analysis With Missing Data. John Wiley
& Sons: New York, 1987.
36. Santos SF, Peixoto AJ. Revisiting the dialysate sodium prescrip-
tion as a tool for better blood pressure and interdialytic weight gain
management in hemodialysis patients. Clin J Am Soc Nephrol 2008;
3:522–530.
37. Zhou YL, Liu J, Ma LJ, Sun F, Shen Y, Huang J, Cui TG. Effects of increas-
ing diffusive sodium removal on blood pressure control in hemodialy-
sis patients with optimal dry weight. Blood Purif 2013; 35:209–215.
38. Suckling RJ, Swift PA, He FJ, Markandu ND, MacGregor GA. Altering
plasma sodium concentration rapidly changes blood pressure during
haemodialysis. Nephrol Dial Transplant 2013; 28:2181–2186.
39. Kim do Y, Kim B, Moon KH, Lee S, Lee DY. Effect of gradually lowering
dialysate sodium concentration on the interdialytic weight gain, blood
pressure, and extracellular water in anuric hemodialysis patients. Renal
Failure 2014; 36:23–27.
40. Shah A, Davenport A. Does a reduction in dialysate sodium improve
blood pressure control in haemodialysis patients? Nephrology 2012;
17:358–363.
41. Rogus JJ, Lin SF, Lacson Jr EK. Efficient Monte Carlo resampling of
continuous data with a dichotomous treatment variable. J Biom Biostat
2012; S7:022. doi:10.4172/2155-6180.S7-022.
42. Port FK, Johnson WJ, Klass DW. Prevention of dialysis disequilibrium
syndrome by use of high sodium concentration in the dialysate. Kidney
Int 1973; 3:327–333.
43. de Paula FM, Peixoto AJ, Pinto LV, Dorigo D, Patricio PJ, Santos SF.
Clinical consequences of an individualized dialysate sodium prescrip-
tion in hemodialysis patients. Kidney Int 2004; 66:1232–1238.
44. Rayner HC, Zepel L, Fuller DS, Morgenstern H, Karaboyas A,
Culleton BF, Mapes DL, Lopes AA, Gillespie BW, Hasegawa T, Saran R,
Tentori F, Hecking M, Pisoni RL, Robinson BM. Recovery time,
quality of life, and mortality in hemodialysis patients: The DOPPS
(Dialysis Outcomes and Practice Patterns Study). Am J Kidney
Dis, E-Published 13 February 2014. http://www.ncbi.nlm.nih.gov/
pubmed/?term=24529994. Accessed 3 October 2014.
45. Agarwal R, Peixoto AJ, Santos SF, Zoccali C. Pre- and postdialysis blood
pressures are imprecise estimates of interdialytic ambulatory blood
pressure. Clin J Am Soc Nephrol 2006; 1:389–398.
46. Agarwal R. Ambulatory blood pressure and cardiovascular events in
chronic kidney disease. Semin Nephrol 2007; 27:538–543.