Epidemiology and Prevention

Hyponatremia, Hypernatremia, and Mortality in Patients

With Chronic Kidney Disease With and Without Congestive
Heart Failure
Csaba P. Kovesdy, MD; Evan H. Lott; Jun Ling Lu, MD; Sandra M. Malakauskas, PhD, MD;
Jennie Z. Ma, PhD; Miklos Z. Molnar, MD, PhD; Kamyar Kalantar-Zadeh, MD, PhD

Background—Hyponatremia is common in patients with conditions such as congestive heart failure and is associated with
increased mortality in hospitalized patients. Congestive heart failure is common in patients with chronic kidney disease,
but the association of serum sodium concentration with mortality in such patients is not well characterized.
Methods and Results—We examined the association of serum sodium concentration with all-cause mortality in a
nationally representative cohort of 655 493 US veterans with non– dialysis-dependent chronic kidney disease (95 961
[15%] of them with congestive heart failure). Associations were examined in time-dependent Cox models with
adjustment for potential confounders. During a median follow-up of 5.5 years, a total of 193 956 patients died (mortality
rate, 62.5/1000 patient-years; 95% confidence interval, 62.2– 62.8). The association of serum sodium level with
mortality was U-shaped, with the lowest mortality seen in patients with sodium level of 140 mEq/L and with both lower
and higher levels showing significant associations with increased mortality. Patients with serum sodium levels of ⬍130,
130 to 135.9, 145.1 to 150, and ⱖ150 mEq/L compared with 136 to 145 mEq/L had multivariable-adjusted mortality
hazard ratios (95% confidence interval) of 1.93 (1.83–2.03), 1.28 (1.26 –1.30), 1.33 (1.28 –1.38), and 1.56 (1.33–1.83)
(P⬍0.001 for all). The associations remained consistent in subgroups of patients with and without congestive heart
failure.
Conclusions—Both lower and higher serum sodium levels are independently associated with higher mortality in
patients with non– dialysis-dependent chronic kidney disease, irrespective of the presence or absence of congestive
heart failure. (Circulation. 2012;125:677-684.)
Key Words: epidemiology 䡲 heart failure 䡲 hypernatremia 䡲 hyponatremia 䡲 kidney

H yponatremia is one of the most common electrolyte

abnormalities that has been described primarily in hos-
pitalized patients; the prevalence of hyponatremia in hospi-
sodium levels have focused on low serum sodium (hypona-
tremia), elevated serum sodium (hypernatremia) has also
been associated with an increase in mortality in hospitalized
talized patients has been reported to be as high as 42% in patients.20
some studies.1,2 Hyponatremia has been associated with
various adverse clinical outcomes such as increased mortal-
Clinical Perspective on p 684
ity,3–20 length of inpatient stay,20,21 gait imbalance and falls,22 Patients with chronic kidney disease (CKD) may be more
rhabdomyolysis,23 and bone fractures.24 –26 Additionally, hy- susceptible to the development of dysnatremias by virtue of
ponatremia has also been linked to significantly increased their diminished ability to maintain water homeostasis in the
healthcare costs.27–29 Most of the studies that examined face of decreasing kidney function. Despite this, other than a
outcomes associated with hyponatremia studied hospitalized single study in hemodialysis patients,19 to our knowledge
patients at single medical centers, and many restricted their there have been no attempts to explore the association of
analyses to patients with various preexisting pathological abnormal serum sodium levels in patients with CKD. We
conditions known to cause hyponatremia, such as congestive examined the association of serum sodium levels measured
heart failure (CHF) and liver cirrhosis. Although most studies repeatedly over time with all-cause mortality in a large,
that examined outcomes associated with abnormal serum nationally representative cohort of US veterans with non–

Received August 30, 2011; accepted December 30, 2011.

From the Division of Nephrology, Salem Veterans Affairs Medical Center, Salem, VA (C.P.K., S.M.M.); Division of Nephrology, University of
Virginia, Charlottesville, VA (C.P.K., S.M.M., J.Z.M.); Veterans Affairs Informatics and Computing Infrastructure, Salt Lake City, UT (E.H.L.); Salem
Research Institute, Salem, VA (J.L.L.); Harold Simmons Center for Chronic Disease Research and Epidemiology, Los Angeles Biomedical Research
Institute at Harbor–UCLA Medical Center, Torrance, CA (M.Z.M., K.K.-Z.); Institute of Pathophysiology, Semmelweis University, Budapest, Hungary
(M.Z.M.); and David Geffen School of Medicine at UCLA, Los Angeles, CA (K.K.-Z.).
Correspondence to Csaba P. Kovesdy, MD, Division of Nephrology, Salem Veterans Affairs Medical Center, 1970 Roanoke Blvd, Salem, VA 24153.
E-mail csaba.kovesdy@va.gov
© 2012 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.111.065391

677
678 Circulation February 7, 2012

Figure 1. Algorithm used to define the

study cohort. eGFR indicates estimated
glomerular filtration rate; ESRD, end-
stage renal disease; and CKD, chronic
kidney disease.

dialysis-dependent CKD. We examined associations with Sociodemographic Characteristics

both mild and moderate to severe hyponatremia and and Comorbidities
hypernatremia. Data on patient age, gender, race, geographic location (Veteran
Integrated Service Network number), and blood pressure were
obtained through the VA Corporate Data Warehouse. Information on
Methods race was complemented with data obtained from Medicare through
the VA-Medicare data merge project.33 All blood pressure values
Cohort Definition available from the time period of October 1, 2004, to September 30,
We used laboratory data on serum creatinine from the Veterans 2009, were recorded and grouped by calendar quarters, and their
Affairs (VA) Decision Support System National Data Extracts quarterly-averaged values were used for analyses. Data on comor-
Laboratory Results file (a VA-wide database containing select bidities (including the presence of CHF, liver disease, malignancies,
laboratory results obtained in the clinical setting)30 to identify and depression) were collected from the VA Inpatient and Outpatient
patients with CKD on the basis of a stable estimated glomerular Medical SAS Datasets34,35 with the use of International Classifica-
filtration rate (eGFR) and the presence of an elevated spot urine tion of Diseases, Ninth Revision diagnostic and procedure codes and
microalbumin/creatinine ratio (for those with eGFR ⱖ60).31 GFR Current Procedural Terminology codes recorded during the time
was estimated from serum creatinine measurements and demo- period of October 1, 2004, to September 30, 2006. These databases
graphic characteristics by the Chronic Kidney Disease Epidemiology contain up to 12 diagnostic and/or procedure codes for every
Collaboration equation.32 The algorithm for cohort definition is inpatient, long-term care, and outpatient VA encounter, as well as
shown in Figure 1. Of a total of 4 381 049 patients with any available non-VA encounters. Prevalent cardiovascular disease was defined as
eGFR between October 1, 2004, and September 30, 2006, we the presence of diagnostic codes for coronary artery disease, angina,
identified 655 493 patients with non– dialysis-dependent CKD and or myocardial infarction or procedure codes for percutaneous coro-
available serum sodium measurements. nary interventions or coronary artery bypass grafting. We calculated
 Kovesdy et al Serum Sodium Level and Outcomes in CKD 679

the Charlson comorbidity index using the Deyo modification for patients with serum sodium levels ⬍115 mEq/L (n⫽76) and ⬎160
administrative data sets, without including kidney disease.36 mEq/L (n⫽61), the spline models were limited to patients with serum
sodium levels of 115 to 160 mEq/L. We also compared patients with
Laboratory Characteristics mild to moderate and severe hyponatremia (serum sodium ⬍130 and
Data on laboratory variables were collected from the time period of 130 –135 mEq/L) and hypernatremia (serum sodium 146 –150 and
October 1, 2004, to September 30, 2009, by using the Decision ⬎150 mEq/L) with those with normal serum sodium (serum sodium
Support System National Data Extracts Laboratory Results file.30 To 136 –145 mEq/L). Because of their smaller numbers, patients with
minimize random variability, all available laboratory values (includ- hypernatremia were analyzed as a single category (serum sodium
ing all serum sodium levels) were grouped by calendar quarters, and ⬎145 mEq/L) in subgroup analyses. To better assess the short-term
their quarterly-averaged values were used in analyses. versus the long-term effects of serum sodium levels on mortality,
time-stratified Cox models were constructed to examine 1-year
Statistical Analyses mortality hazard ratios associated with baseline hyponatremia and
Descriptive analyses were performed, and skewed variables were hypernatremia (defined by serum sodium levels at the cohort entry
log-transformed. Because of the large sample size, traditional statis- and maintained constant throughout the examined time periods) and
tical testing of differences in baseline characteristics was statistically with time-varying hyponatremia and hypernatremia (defined on the
significant for all variables; hence, the significance of differences basis of repetitive quarterly serum sodium measurements in each
was established on the basis of values that we deemed to be examined year) in the first, second, third, fourth, and fifth years of
biologically meaningful differences. Data points were missing for follow-up, conditional on surviving to the beginning of the examined
race (1.4%), blood pressure (17.5%), serum albumin (13.0%), (first, second, etc) year.42
hemoglobin (13.2%), white blood cell count (16.1%), aspartate The association of serum sodium with mortality was examined
aminotransferase (7.4%), alanine aminotransferase (5.3%), total separately in subgroups of patients categorized by CKD stage, by
bilirubin (9.4%), alkaline phosphatase (10.3%), and blood glucose key sociodemographic characteristics, by presence or absence of key
(0.5%). There were a total of 509 906 patients (78% of the total study comorbid conditions, and by their levels of relevant laboratory
population) with complete data available for the fully adjusted values. Sensitivity analyses were performed with the use of imputed
multivariable analyses. Compared with patients with missing data, values of independent variables and serum sodium levels corrected
patients with complete data were of similar age (73.8⫾9.6 versus for serum glucose level. Statistical analyses were performed with the
74.0⫾10.1 years), gender (2.7% versus 3.1% female), and race (87% use of STATA MP version 11 (STATA Corporation, College
versus 91% white, 10% versus 7% black) and had similar prevalence Station, TX). The study protocol was approved by the Research and
of diabetes mellitus (44% versus 41%), cardiovascular disease (44% Development Committee at the Salem VA Medical Center.
versus 39%), and CHF (15% versus 12%). Missing values were not
imputed in primary analyses and were substituted with the use of Results
multiple imputation procedures in sensitivity analyses. Missing The mean⫾SD age of the cohort at baseline was 73.9⫾9.8
values were replaced by multiple imputations with a multivariate years; 87% and 9% of patients were white and black,
normal regression method with data augmentation by an iterative
Markov chain Monte Carlo procedure37,38 in STATA’s “mi” com- respectively; and the mean eGFR was 50.2⫾14.1 mL/min per
mand suite. Ten imputed data sets were generated; primary analyses 1.73 m2. The mean⫾SD baseline serum sodium was 140⫾3
were performed on each imputed data set, and the combination rules mEq/L. At baseline, 85 855 patients (13.5%) had hyponatre-
of Rubin39 were used to form 1 set of results. mia (serum sodium ⬍136 mEq/L), and 13 289 (2%) had
The start of the follow-up period was the date of the first available
hypernatremia (serum sodium ⬎145 mEq/L); during the
serum sodium measurement after October 1, 2004. Patients were
followed until death or were censored at the date of the last entire duration of follow-up, 169 158 patients (26%) had at
healthcare or administrative VA encounter, as documented in the VA least 1 episode of hyponatremia, and 45 666 (7%) had at least
Vital Status Files. The VA Vital Status Files is a registry containing 1 episode of hypernatremia. Baseline characteristics in pa-
dates of death or last medical/administrative encounter from all tients categorized by their baseline serum sodium levels are
available sources in the VA system (Beneficiary Identification
Records Locator Subsystem, Patient Treatment File, Medicare, and shown in Table 1. Patients with hyponatremia were younger;
Social Security Administration). The sensitivity and specificity of were more likely to be diabetic and to have CHF, liver
the Vital Status Files with the National Death Index used as gold disease, and depression; had a higher eGFR, blood glucose,
standard were shown to be very high (98.3% and 99.8%, respec- and white blood cell count; and had a lower serum albumin
tively).40 For 2956 patients (0.005%) with missing Vital Status Files and blood hemoglobin. Patients with hypernatremia, on the
data, the date of the last available laboratory measurement was used
as the censoring date. The association of serum sodium level with other hand, were older and had a lower eGFR, serum total
all-cause mortality was examined in time-dependent Cox models, bilirubin, and blood glucose.
with adjustment for potential confounders. Variables were included
in multivariable models if they could be considered confounders41 on Mortality
the basis of theoretical considerations and after examination of A total of 193 956 patients died (mortality rate, 62.5/1000
baseline associations with serum sodium. Associations were exam-
ined sequentially in models with incremental multivariable adjust- patient-years; 95% confidence interval, 62.2– 62.8) during a
ments: unadjusted (model 1); age, gender, race, and geographic median follow-up of 5.5 years. The number of deaths in
location (model 2); model 2 plus comorbid conditions (diabetes patients with different serum sodium levels were 1773 (⬍130
mellitus, atherosclerotic cardiovascular disease, CHF, liver disease, mEq/L); 30 199 (130 –135.9 mEq/L); 158 103 (136 –144.9
malignancy, depression, and Charlson comorbidity index) (model 3);
mEq/L); 3680 (145–149.9 mEq/L); and 201 (ⱖ150 mEq/L).
and model 3 plus systolic blood pressure, eGFR, serum albumin,
alkaline phosphatase, aspartate aminotransferase, alanine amino- The association of serum sodium with mortality was
transferase, total bilirubin, blood hemoglobin, glucose, and white U-shaped, with both lower and higher serum sodium showing
blood cell count (model 4). Variables that were measured repeatedly a significant association with higher mortality even after
during follow-up (serum sodium, blood pressure, and all other multivariable adjustment (Figure 2). Patients with serum
laboratory covariates) were handled as time-dependent variables in
Cox models. We hypothesized that the association of serum sodium sodium levels of ⬍130, 130 to 135.9, 145.1 to 149.9, and
with mortality will be nonlinear; hence, we examined sodium by ⱖ150 mEq/L compared with 136 to 145 mEq/L had unad-
using restricted cubic splines. Because of the small number of justed mortality hazard ratios (95% confidence interval) of
680 Circulation February 7, 2012

Table 1. Baseline Characteristics of Individuals Stratified by Baseline Serum Sodium Level

Serum Sodium, mEq/L

⬍130 (n⫽2729) 130 –135.9 (n⫽83 126) 136 –144.9 (n⫽556 349) 145–149.9 (n⫽12 807) ⱖ150 (n⫽482)
Age, y 71.5⫾11.3 71.7⫾10.7 74.1⫾9.6 75.5⫾8.9 76.1⫾8.5
Race
White 2346 (87) 70 905 (87) 482 718 (88) 11 245 (89) 423 (89)
Black 217 (8) 8206 (10) 51 704 (9) 1163 (9) 40 (8)
Hispanic 49 (2) 998 (1) 7107 (1) 146 (1) 9 (2)
Other 76 (3) 1566 (2) 7694 (1) 136 (1) 5 (1)
Gender (male) 2631 (96) 80 637 (97) 541 195 (97) 12 430 (97) 474 (98)
Diabetes mellitus 1302 (48) 45 761 (55) 230 989 (42) 4873 (38) 185 (38)
Atherosclerotic CVD 1112 (41) 34 868 (42) 237 429 (43) 5778 (45) 221 (46)
CHF 556 (20) 14 076 (17) 79 374 (14) 1880 (15) 75 (16)
Liver disease 101 (4) 1257 (2) 3390 (0.6) 51 (0.4) 1 (0.2)
Malignancy 454 (17) 13 949 (17) 98 068 (18) 2445 (19) 82 (17)
Depression 205 (8) 5740 (7) 30 295 (5) 692 (5) 24 (5)
Comorbidity index 3 (2–5) 3 (2–5) 3 (2–5) 3 (2–5) 3 (2–4)
SBP, mm Hg 137⫾20 137⫾18 137⫾18 138⫾18 135⫾19
DBP, mm Hg 72⫾11 72⫾11 72⫾11 72⫾11 70⫾12
2
eGFR, mL/min per 1.73 m 55.2⫾19.3 53.1⫾17.3 49.8⫾13.5 47.1⫾12.6 47.0⫾13.1
Serum albumin, g/dL 3.8⫾0.5 3.9⫾0.5 4.0⫾0.4 4.1⫾0.4 4.2⫾0.5
Total cholesterol, mg/dL 173⫾67 178⫾46 173⫾39 171⫾38 170⫾37
Serum calcium, mg/dL 9.1⫾0.5 9.2⫾0.5 9.3⫾0.5 9.5⫾0.5 9.5⫾0.6
Serum AST, U/L 24 (19–31) 22 (18–28) 22 (19–27) 24 (20–29) 25 (21–30)
Serum ALT, U/L 22 (16–32) 22 (16–31) 22 (16–30) 25 (18–32) 26 (19–33)
Serum total bilirubin, mg/dL 0.7 (0.5–0.9) 0.7 (0.5–0.9) 0.7 (0.5–0.9) 0.6 (0.4–0.7) 0.5 (0.4–0.7)
Serum ALP, U/L 91⫾46 81⫾38 79⫾32 85⫾30 87⫾31
Serum bicarbonate, mEq/L 26.0⫾3.0 26.6⫾2.8 27.4⫾2.9 27.9⫾3.3 28.0⫾4.2
Blood hemoglobin, g/dL 13.1⫾1.9 13.7⫾1.8 13.9⫾1.7 13.9⫾1.7 14.0⫾1.7
Blood WBC, 1000/mm3 7.9⫾3.0 7.8⫾3.6 7.3⫾4.1 7.3⫾3.4 7.4⫾2.2
Blood glucose, mg/dL 166⫾126 151⫾48 119⫾40 109⫾28 110⫾32
Data are presented as mean⫾SD, No. (% of total), or median (quartiles 1–3). CVD indicates cardiovascular disease; CHF, congestive heart failure; SBP, systolic
blood pressure; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ALP,
alkaline phosphatase; and WBC, white blood cell count. All P values for comparing differences between categories were statistically significant.

2.49 (2.38 –2.61), 1.43 (1.41–1.44), 1.40 (1.34 –1.35), and associated with hyponatremia in patients with different stages
2.17 (1.89 –2.49) (P⬍0.001 for all); after full multivariable of CKD (Figure 4A); however, mortality associated with
adjustment, the hazard ratios (95% confidence interval) for hypernatremia appeared to be relatively lower in patients with
the same groups were 1.93 (1.83–2.03), 1.28 (1.26 –1.30), more advanced stages of CKD (Figure 4B).
1.33 (1.28 –1.38), and 1.56 (1.33–1.83) (P⬍0.001 for all) Results of analyses in which imputed values for missing
(Figure 3). In time-stratified Cox models concomitantly variables were used yielded similar results: The multivariable
assessing 1-year mortality rates associated with both baseline adjusted hazard ratios (95% confidence interval) in patients
and time-varying serum sodium categories, time-varying with serum sodium levels of ⬍130, 130 to 135.9, 145.1 to
hyponatremia and hypernatremia both showed stronger asso- 149.9, and ⱖ150 mEq/L compared with 136 to 145 mEq/L
ciations with mortality compared with baseline hyponatremia were 1.94 (1.85–2.03), 1.28 (1.27–1.30), 1.29 (1.25–1.34),
and hypernatremia, which displayed weak or nonsignificant and 1.58 (1.38 –1.82) (P⬍0.001 for all). The results of
associations (Table 2). analyses examining serum sodium levels that were adjusted
The association of lower and higher serum sodium with for blood glucose levels were also not different from the
mortality was present in all examined subgroups, including results detailed above (data not shown).
patients with and without CHF, liver disease, malignancy,
and depression; patients with high and low levels of the Discussion
composite Charlson comorbidity index; and patients with We describe an association between abnormally decreased
normal or elevated serum levels of hepatic enzymes (Figure and elevated levels of serum sodium and higher all-cause
4). There was no linear trend in the mortality hazard ratios mortality in a large, nationally representative group of US
 Kovesdy et al Serum Sodium Level and Outcomes in CKD 681

rather than chronic (long-term) risk factors for mortality.

Severity of kidney disease did not appear to affect the
mortality associated with hyponatremia, but patients with
more advanced CKD displayed a relatively lower mortality
associated with hypernatremia compared with patients with
less severe stages of CKD.
Studies examining the predictive value of serum sodium
level have concentrated largely on hyponatremia, have exam-
ined mostly patients who were hospitalized, and were usually
derived from data obtained from single medical centers. Such
studies have described associations of hyponatremia with a
variety of adverse outcomes including all-cause mortality,3–20
length of inpatient stay,20,21 gait imbalance and falls,22 rhab-
domyolysis,23 bone fractures,24 –26 and higher hospitalization
Figure 2. Multivariable-adjusted log hazard ratios (95% confi- costs.27–29 Some of the studies examined unselected groups of
dence intervals) of all-cause mortality associated with serum patients,18,20 but others focused on groups with some under-
sodium levels in a time-dependent Cox model with the use of
restricted cubic splines, adjusted for age, gender, race, geo-
lying comorbid condition such as CHF,7,11,14 cardiovascular
graphic location, diabetes mellitus, atherosclerotic cardiovascu- disease,9,10 or liver disease.4,13 Irrespective of the setting or
lar disease, congestive heart failure, liver disease, malignancy, the patients included, all studies have found that hyponatre-
depression, Charlson comorbidity index, systolic blood pres- mia is associated with an increased risk of the studied end
sure, estimated glomerular filtration rate, serum albumin, alkaline
phosphatase, aspartate and alanine aminotransferase, total bili- points. Similar associations were reported for hypernatremia
rubin, blood hemoglobin, glucose, and white blood cell count. as well,20 although in general this abnormality has been
The bars represent the number of deaths in patients with serum underemphasized.
sodium levels grouped in increments of 5 mEq/L from 115 to
160 mEq/L, on a logarithmic scale.
Ours is the first study that examined patients with non–
dialysis-dependent CKD and the first to provide data that are
nationally representative for the United States. The unique-
veterans with non– dialysis-dependent CKD. These associa- ness of the CKD population is that kidney disease affects the
tions were linearly proportional to the severity of the under- organ responsible for maintaining water homeostasis, and as
lying hyponatremia and hypernatremia and were independent such it is possible that both the prevalence of dysnatremias
of comorbid conditions such as CHF or liver disease. Both and their clinical consequences could be magnified. Although
lower and higher serum sodium were associated more we reported a relatively high prevalence and incidence of
strongly with mortality when they were modeled in a time- hyponatremia (13% of the patients in our study had hypona-
dependent manner, with weak or no effects associated with tremia at baseline, and twice as many had at least 1 episode
baseline levels of the same factors, suggesting that both of hyponatremia during follow-up), the lack of information
hyponatremia and hypernatremia represent acute (short-term) about the general population prevents us from determining
whether CKD results in an increased incidence or prevalence
of dysnatremias. In regard to outcomes, we did not find
differences in the association of hyponatremia with mortality
in patients with different severities of CKD, but hyperna-
tremia appeared to predict less severe outcomes in patients
with more advanced stages of CKD. This latter observation
could be the result of end-organ adaptation to a state of
increased extracellular osmolality in patients with advanced
CKD who experience a gradual accumulation of various
uremic solutes with advancing severity of kidney disease.
One implication of our results is that they establish
hyponatremia and hypernatremia as robust outcome predic-
Figure 3. Unadjusted and multivariable-adjusted hazard ratios tors in patients with all stages of CKD. On the other hand, it
(95% confidence intervals) of all-cause mortality associated with is unclear whether these abnormalities should be considered
various levels of serum sodium in time-dependent Cox models.
treatment targets in these patients. Both hyponatremia and
The groups with serum sodium 136 to 145 mEq/L served as
referent. Models represent unadjusted association (model 1) and hypernatremia can have direct adverse effects on the function
associations after adjustment for age, gender, race, and geo- of various organs, most notably on the central nervous
graphic location (model 2); model 2 variables plus diabetes mel- system.43– 46 This underlying pathophysiology could serve as
litus, atherosclerotic cardiovascular disease, congestive heart
failure, liver disease, malignancy, depression, and Charlson a potential explanation of why an abnormal serum sodium
comorbidity index (model 3); and model 3 variables plus systolic level is associated with increased mortality; our results
blood pressure, estimated glomerular filtration rate, serum albu- indicating a more marked association with short-term rather
min, alkaline phosphatase, aspartate and alanine aminotransfer- than long-term mortality also support this hypothesis. Be-
ase, total bilirubin, blood hemoglobin, glucose, and white blood
cell count (model 4). All comparisons were significant at the cause abnormal water homeostasis usually develops as a
P⬍0.001 level. result of another underlying pathology, it is also possible that
682 Circulation February 7, 2012

Table 2. One-Year Unadjusted Mortality Hazard Ratios (95% Confidence Intervals) Associated With Baseline and With
Time-Dependent Serum Sodium Levels of <130, 130 –135.9, and >145 mEq/L Compared With 136 –145 mEq/L in the First, Second,
Third, Fourth, and Fifth Years of Follow-Up, Conditional on Surviving to the Beginning of the Examined Year
Year of Follow-Up

Serum Sodium, mEq/L 1 2 3 4 5

Baseline ⬍130 1.14 (0.93–1.40) 0.99 (0.85–1.17) 1.04 (0.90 –1.21) 1.30 (1.11–1.51) 1.13 (0.95–1.33)
Time-dependent ⬍130 2.00 (1.67–2.38) 2.04 (1.81–2.30) 2.05 (1.82–2.30) 1.63 (1.43–1.86) 1.32 (1.15–1.51)
Baseline 130–135.9 1.03 (0.98–1.09) 1.05 (1.01–1.09) 1.06 (1.03–1.10) 1.04 (1.01–1.08) 1.06 (1.02–1.09)
Time-dependent 130–135.9 1.34 (1.27–1.41) 1.28 (1.23–1.32) 1.30 (1.26–1.34) 1.27 (1.23–1.32) 1.17 (1.13–1.21)
Baseline ⬎145 1.09 (0.94–1.27) 1.18 (1.08–1.29) 1.17 (1.08–1.27) 1.12 (1.02–1.22) 1.02 (0.93–1.11)
Time-dependent ⬎145 1.40 (1.20–1.64) 1.40 (1.28–1.53) 1.37 (1.27–1.49) 1.30 (1.19–1.42) 1.20 (1.10–1.31)
Hazard ratios are estimated from Cox models that included both baseline and time-dependent serum sodium categories and were adjusted for age, gender, race,
geographic location, diabetes mellitus, atherosclerotic cardiovascular disease, congestive heart failure, liver disease, malignancy, depression, the Charlson comorbidity
index, systolic blood pressure, estimated glomerular filtration rate, serum albumin, alkaline phosphatase, aspartate and alanine aminotransferase, total bilirubin, blood
hemoglobin, glucose, and white blood cell count. Baseline serum sodium categories were established on the basis of measurements in the first 3 months of cohort
participation for all models and were kept constant throughout follow-up. Time-dependent serum sodium categories were established with the use of repetitive
quarterly serum sodium levels measured throughout follow-up.

hyponatremia and hypernatremia are merely surrogate mark- hemodialysis patients enrolled in the Hemodialysis (HEMO)
ers of more severe disease states. Similar to other studies, we study also reported a significant association of hyponatremia
did not detect any effect modification by known disease states with mortality, even though in patients with anuric dialysis
that affect serum sodium level, which makes it more likely the development of low serum sodium is unrelated to the
that an abnormal serum sodium level has an independent stimulation of arginine vasopressin by underlying comorbidi-
effect on survival. Furthermore, a study of maintenance ties.19 Arguing against a causal effect of hyponatremia was a

Figure 4. Forest plot of the multivariable-adjusted natural log-transformed mortality hazard ratios (95% confidence intervals) associated
with mild (130 –135.9 mEq/L) and moderate to severe (⬍130 mEq/L) hyponatremia (A) and with hypernatremia (serum sodium ⬎145
mEq/L) (B) in various prespecified subgroups of patients. Groups with normal (136 –145 mEq/L) serum sodium levels served as referent.
Estimates are from time-dependent Cox models adjusted for age, gender, race, geographic location, diabetes mellitus, atherosclerotic
cardiovascular disease, congestive heart failure, liver disease, malignancy, depression, Charlson comorbidity index, systolic blood pres-
sure, estimated glomerular filtration rate, serum albumin, alkaline phosphatase (AlkPhos), aspartate aminotransferase (AST), alanine
aminotransferase (ALT), total bilirubin, blood hemoglobin, glucose, and white blood cell count. CKD indicates chronic kidney disease.
 Kovesdy et al Serum Sodium Level and Outcomes in CKD 683

recent study in hospitalized patients in whom the authors Conclusions

could link fatalities associated with severe hyponatremia Both hyponatremia and hypernatremia are associated with
(⬍120 mEq/L) to more severe underlying disease processes increased all-cause mortality in patients with non– dialysis-
rather than the hyponatremia itself.47 Nevertheless, observa- dependent CKD. This association is independent of comorbid
tional studies cannot completely overcome the problem of conditions and severity of kidney disease. Abnormal serum
residual confounding, which can be better addressed by sodium levels can be used as predictors of outcomes in this
randomized controlled trials. One such large trial in patients patient population and can be considered treatment targets
that examined the effects of the vasopressin V2-receptor that need to be tested in clinical trials.
antagonist tolvaptan on mortality in patients with CHF did
not detect a significant benefit on mortality,48 but more Acknowledgments
interventional studies are needed to determine whether other Parts of this material were presented at the American Society of
patient populations, other treatment regimens, or patients Nephrology Renal Week 2011; November 8 to 13, 2011; Phila-
delphia, PA.
with different severities of hyponatremia might show differ-
ent outcomes. Sources of Funding
Our study is notable for its very large sample size and event This study was supported by grant 1R01DK078106-01 from the
numbers and the fact that it is representative of the entire National Institute of Diabetes and Digestive and Kidney Diseases to
geographic United States. Our study also has a number of Drs Kovesdy and Kalantar-Zadeh and by resources from the Depart-
ment of Veterans Affairs. The funding sources had no role in the
limitations that should be considered when the results are
design of the study, the data analysis, or the writing of the
interpreted. Our study population consisted mostly of male manuscript. Dr Kovesdy, Evan H. Lott, and Dr Malakauskas are
patients; hence, the results may not apply to female patients. employees of the Department of Veterans Affairs. Opinions ex-
We used data obtained during the course of clinical practice, pressed in this article are those of the authors and do not necessarily
and therefore selection bias is possible. However, the key represent the opinion of the Department of Veterans Affairs.
laboratory variables used for cohort definition (serum creat-
inine and sodium) are part of routine panels that are measured
Disclosures
Dr Kalantar-Zadeh has received honoraria from Otsuka Pharmaceu-
in most patients receiving healthcare, and therefore it is ticals. The other authors report no conflicts.
unlikely that a significant proportion of actively enrolled
veterans would have been excluded. We defined CKD using References
the Chronic Kidney Disease Epidemiology Collaboration 1. Upadhyay A, Jaber BL, Madias NE. Incidence and prevalence of hypo-
equation because it is more accurate than other estimating natremia. Am J Med. 2006;119:S30 –S35.
2. Upadhyay A, Jaber BL, Madias NE. Epidemiology of hyponatremia.
equations (such as the Modification of Diet in Renal Disease Semin Nephrol. 2009;29:227–238.
equation) in patients with normal and mildly decreased GFR. 3. Anderson RJ, Chung HM, Kluge R, Schrier RW. Hyponatremia: a pro-
The Chronic Kidney Disease Epidemiology Collaboration spective analysis of its epidemiology and the pathogenetic role of vaso-
pressin. Ann Intern Med. 1985;102:164 –168.
equation was, however, meant to be used with serum creati- 4. Borroni G, Maggi A, Sangiovanni A, Cazzaniga M, Salerno F. Clinical
nine measured by the isotope dilution mass spectrometry– relevance of hyponatraemia for the hospital outcome of cirrhotic patients.
traceable method, which was not ubiquitous at the time when Dig Liver Dis. 2000;32:605– 610.
5. Chung HM, Kluge R, Schrier RW, Anderson RJ. Postoperative hypona-
our cohort was defined (2005–2006), and hence the accuracy
tremia: a prospective study. Arch Intern Med. 1986;146:333–336.
of the staging of CKD in our cohort is unclear. The associ- 6. Clayton JA, Le Jeune IR, Hall IP. Severe hyponatraemia in medical
ations between serum sodium and mortality did not change, in-patients: aetiology, assessment and outcome. QJM. 2006;99:505–511.
however, if we used the Modification of Diet in Renal 7. Gheorghiade M, Abraham WT, Albert NM, Gattis SW, Greenberg BH,
O’Connor CM, She L, Yancy CW, Young J, Fonarow GC. Relationship
Disease equation to estimate GFR in our cohort participants between admission serum sodium concentration and clinical outcomes in
(data not shown). We did not collect information about patients hospitalized for heart failure: an analysis from the
hospitalizations, and therefore we cannot determine whether OPTIMIZE-HF registry. Eur Heart J. 2007;28:980 –988.
8. Gill G, Huda B, Boyd A, Skagen K, Wile D, Watson I, Van Heyningen
the mortality associated with hyponatremia or hypernatremia C. Characteristics and mortality of severe hyponatraemia: a
occurred in the context of hospitalizations. We adjusted our hospital-based study. Clin Endocrinol (Oxf). 2006;65:246 –249.
analyses for a significant number of potential confounders, 9. Goldberg A, Hammerman H, Petcherski S, Zdorovyak A, Yalonetsky S,
Kapeliovich M, Agmon Y, Markiewicz W, Aronson D. Prognostic
but we cannot rule out the presence of residual confounding.
importance of hyponatremia in acute ST-elevation myocardial infarction.
We used diagnostic codes to define comorbid conditions that Am J Med. 2004;117:242–248.
could act as confounders in the association of serum sodium 10. Goldberg A, Hammerman H, Petcherski S, Nassar M, Zdorovyak A,
with mortality, which may have resulted in underestimation Yalonetsky S, Kapeliovich M, Agmon Y, Beyar R, Markiewicz W,
Aronson D. Hyponatremia and long-term mortality in survivors of acute
of their prevalence. By including laboratory variables reflect- ST-elevation myocardial infarction. Arch Intern Med. 2006;166:781–786.
ing abnormal liver function and/or structure, we were able to 11. Klein L, O’Connor CM, Leimberger JD, Gattis-Stough W, Pina IL, Felker
alleviate this concern as far as the confounding imparted by GM, Adams KF Jr, Califf RM, Gheorghiade M. Lower serum sodium is
associated with increased short-term mortality in hospitalized patients
liver disease, but we did not have similar data to assess the with worsening heart failure: results from the Outcomes of a Prospective
presence and/or severity of the other relevant comorbid Trial of Intravenous Milrinone for Exacerbations of Chronic Heart Failure
conditions. We had relatively few patients with extremely (OPTIME-CHF) study. Circulation. 2005;111:2454 –2460.
high or extremely low serum sodium levels, and therefore our 12. Lee CT, Guo HR, Chen JB. Hyponatremia in the emergency department.
Am J Emerg Med. 2000;18:264 –268.
ability to characterize outcomes associated with very severe 13. Londono MC, Guevara M, Rimola A, Navasa M, Taura P, Mas A,
degrees of hyponatremia and hypernatremia may be limited. Garcia-Valdecasas JC, Arroyo V, Gines P. Hyponatremia impairs early
684 Circulation February 7, 2012

posttransplantation outcome in patients with cirrhosis undergoing liver Department of Veterans Affairs, Health Services Research and Devel-
transplantation. Gastroenterology. 2006;130:1135–1143. opment Service, VA Information Resource Center; 2009.
14. Mohammed AA, van Kimmenade RR, Richards M, Bayes-Genis A, Pinto 31. K/DOQI clinical practice guidelines for chronic kidney disease: eval-
Y, Moore SA, Januzzi JL Jr. Hyponatremia, natriuretic peptides, and uation, classification, and stratification. Am J Kidney Dis. 2002;39:
outcomes in acutely decompensated heart failure: results from the Inter- S1–S266.
national Collaborative of NT-proBNP Study. Circ Heart Fail. 2010;3: 32. Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF III, Feldman
354 –361. HI, Kusek JW, Eggers P, Van LF, Greene T, Coresh J. A new equation
15. Sterns RH. Severe symptomatic hyponatremia: treatment and outcome: a to estimate glomerular filtration rate. Ann Intern Med. 2009;150:
study of 64 cases. Ann Intern Med. 1987;107:656 – 664. 604 – 612.
16. Terzian C, Frye EB, Piotrowski ZH. Admission hyponatremia in the 33. US Department of Veterans Affairs. VA Information Resource Center
elderly: factors influencing prognosis. J Gen Intern Med. 1994;9:89 –91. Data Quality Update: Race. Hines, IL: US Dept of Veterans Affairs, VA
17. Tierney WM, Martin DK, Greenlee MC, Zerbe RL, McDonald CJ. The Information Resource Center; 2009.
prognosis of hyponatremia at hospital admission. J Gen Intern Med. 34. VIReC Research User Guide: VHA Medical SAS Inpatient Datasets FY
1986;1:380 –385. 2006. Hines, IL: US Dept of Veterans Affairs, VA Information Resource
18. Waikar SS, Mount DB, Curhan GC. Mortality after hospitalization with Center; 2007.
mild, moderate, and severe hyponatremia. Am J Med. 2009;122:857– 865. 35. VIReC Research User Guide: VHA Medical SAS Outpatient Datasets FY
19. Waikar SS, Curhan GC, Brunelli SM. Mortality associated with low 2006. Hines, IL: US Dept of Veterans Affairs, VA Information Resource
serum sodium concentration in maintenance hemodialysis. Am J Med. Center; 2007.
2011;124:77– 84. 36. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index
20. Wald R, Jaber BL, Price LL, Upadhyay A, Madias NE. Impact of for use with ICD-9-CM administrative databases. J Clin Epidemiol.
hospital-associated hyponatremia on selected outcomes. Arch Intern Med. 1992;45:613– 619.
37. Li KH. Imputation using Markov chains. J Stat Comput Simul. 1988;30:
2010;170:294 –302.
57–79.
21. Sherlock M, O’Sullivan E, Agha A, Behan LA, Rawluk D, Brennan P,
38. Tanner MA, Wong WH. The calculation of posterior distributions by data
Tormey W, Thompson CJ. The incidence and pathophysiology of
augmentation. J Am Stat Assoc. 1987;82:528 –550.
hyponatraemia after subarachnoid haemorrhage. Clin Endocrinol (Oxf).
39. Rubin DB. Multiple Imputation for Nonresponse in Surveys. New York,
2006;64:250 –254.
NY: J Wiley & Sons; 1987.
22. Renneboog B, Musch W, Vandemergel X, Manto MU, Decaux G. Mild
40. Arnold N, Sohn M, Maynard C, Hynes DM. VIReC Technical Report 2:
chronic hyponatremia is associated with falls, unsteadiness, and attention
VA-NDI Mortality Data Merge Project. Hines, IL: VA Information Resource
deficits. Am J Med. 2006;119:71–78.
Center; 2006.
23. Morita S, Inokuchi S, Yamamoto R, Inoue S, Tamura K, Ohama S, 41. Thadhani R, Tonelli M. Cohort studies: marching forward. Clin J Am Soc
Nakagawa Y, Yamamoto I. Risk factors for rhabdomyolysis in self- Nephrol. 2006;1:1117–1123.
induced water intoxication (SIWI) patients. J Emerg Med. 2010;38: 42. Dekker FW, de Mutsert R, van Dijk PC, Zoccali C, Jager KJ. Survival
293–296. analysis: time-dependent effects and time-varying risk factors. Kidney Int.
24. Kinsella S, Moran S, Sullivan MO, Molloy MG, Eustace JA. Hyponatre- 2008;74:994 –997.
mia independent of osteoporosis is associated with fracture occurrence. 43. Arieff AI, Guisado R. Effects on the central nervous system of hyperna-
Clin J Am Soc Nephrol. 2010;5:275– 80. tremic and hyponatremic states. Kidney Int. 1976;10:104 –116.
25. Gankam KF, Andres C, Sattar L, Melot C, Decaux G. Mild hyponatremia 44. Mount DB. The brain in hyponatremia: both culprit and victim. Semin
and risk of fracture in the ambulatory elderly. QJM. 2008;101:583–588. Nephrol. 2009;29:196 –215.
26. Ayus JC, Moritz ML. Bone disease as a new complication of hyponatre- 45. Sterns RH, Silver SM. Brain volume regulation in response to hypo-
mia: moving beyond brain injury. Clin J Am Soc Nephrol. 2010;5: osmolality and its correction. Am J Med. 2006;119:S12–S16.
167–168. 46. Kalantar-Zadeh K, Nguyen MK, Chang R, Kurtz I. Fatal hyponatremia in
27. Callahan MA, Do HT, Caplan DW, Yoon-Flannery K. Economic impact a young woman after ecstasy ingestion. Nat Clin Pract Nephrol. 2006;2:
of hyponatremia in hospitalized patients: a retrospective cohort study. 283–288, quiz.
Postgrad Med. 2009;121:186 –191. 47. Chawla A, Sterns RH, Nigwekar SU, Cappuccio JD. Mortality and serum
28. Hirth RA, Messana JM. What does a serum sodium cost? J Am Soc sodium: do patients die from or with hyponatremia? Clin J Am Soc
Nephrol. 2008;19:654 – 655. Nephrol. 2011;6:960 –965.
29. Shea AM, Hammill BG, Curtis LH, Szczech LA, Schulman KA. Medical 48. Konstam MA, Gheorghiade M, Burnett JC Jr, Grinfeld L, Maggioni AP,
costs of abnormal serum sodium levels. J Am Soc Nephrol. 2008;19: Swedberg K, Udelson JE, Zannad F, Cook T, Ouyang J, Zimmer C,
764 –770. Orlandi C. Effects of oral tolvaptan in patients hospitalized for worsening
30. VIReC Research User Guide: Veterans Health Administration Decision heart failure: the EVEREST Outcome Trial. JAMA. 2007;297:
Support System Clinical National Data Extracts. 2nd ed. Hines, IL: US 1319 –1331.

CLINICAL PERSPECTIVE
We examined associations of serum sodium levels with all-cause mortality in 655 493 US veterans with chronic kidney
disease, 15% of whom suffered from congestive heart failure. During a median follow-up of 5.5 years, 169 158 patients
(26%) had at least 1 episode of hyponatremia, and 45 666 (7%) had at least 1 episode of hypernatremia. Both hyponatremia
and hypernatremia were associated with significantly higher short-term mortality, independent of comorbidities such as
congestive heart failure or liver disease. Severity of kidney disease did not affect the association of hyponatremia with
mortality, but patients with more advanced chronic kidney disease appeared to tolerate hypernatremia relatively better.
These observations extend the findings of earlier studies regarding the association of hyponatremia and hypernatremia with
increased mortality to a patient population with chronic kidney disease, which is known to suffer from a high burden of
comorbidities and a very high cardiovascular mortality rate. Even though the results of this study cannot prove that
hyponatremia and hypernatremia are causes of higher mortality, they emphasize the important predictive value of serum
sodium levels in patients with chronic kidney disease and establish this patient population as a potential target for future
clinical trials testing interventions to correct abnormal serum sodium levels.