JBMR

ORIGINAL ARTICLE

Mild Hyponatremia as a Risk Factor for Fractures:

The Rotterdam Study
Ewout J Hoorn, 1 Fernando Rivadeneira, 1,2 Joyce BJ van Meurs, 1 Gijsbertus Ziere, 1,2 Bruno HCh Stricker, 2
Albert Hofman , 2 Huibert AP Pols , 1,2 Robert Zietse, 1 Andre G Uitterlinden , 1,2 and M Carola Zillikens1
1
2

Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands

Department of Epidemiology, Erasmus Medical Center, Rotterdam, The Netherlands

ABSTRACT
Recent studies suggest that mild hyponatremia is associated with fractures, but prospective studies are lacking. We studied whether
hyponatremia is associated with fractures, falls, and/or bone mineral density (BMD). A total of 5208 elderly subjects with serum sodium
assessed at baseline were included from the prospective population-based Rotterdam Study. The following data were analyzed: BMD,
vertebral fractures (mean follow-up 6.4 years), nonvertebral fractures (7.4 years), recent falls, comorbidity, medication, and mortality.
Hyponatremia was detected in 399 subjects (7.7%, 133.4
2.0 mmol/L). Subjects with hyponatremia were older (73.5
10.3 years versus
70.0
9.0 years, p < .001), had more recent falls (23.8% versus16.4%, p < .01), higher type 2 diabetes mellitus prevalence (22.2% versus
10.3%, p < .001), and more often used diuretics (31.1% versus 15.0%, p < .001). Hyponatremia was not associated with lower BMD but
was associated with increased risk of incident nonvertebral fractures [hazard ratio (HR) 1.39, 95% confidence interval (CI) 1.111.73,
p .004] after adjustment for age, sex, and body mass index. Further adjustments for disability index, use of diuretics, use of
psycholeptics, recent falls, and diabetes did not modify results. In the fully adjusted model, subjects with hyponatremia also had
increased risk of vertebral fractures at baseline [odds ratio (OR) 1.78, 95% CI 1.043.06, p .037] but not at follow-up. Finally, all-cause
mortality was higher in subjects with hyponatremia (HR 1.21, 95% CI 1.031.43, p .022). It is concluded that mild hyponatremia in the
elderly is associated with an increased risk of vertebral fractures and incident nonvertebral fractures but not with BMD. Increased fracture
risk in hyponatremia also was independent of recent falls, pointing toward a possible effect on bone quality. 2011 American Society for
Bone and Mineral Research.
KEY WORDS: BONE MINERAL DENSITY; DIURETICS; FALLS; MORTALITY; OSTEOPOROSIS

Introduction

yponatremia, usually defined as a serum sodium concentration of less than 136 mmol/L, is the most common
electrolyte disorder in hospitalized patients.(1) Severe complications of hyponatremia have long been recognized and include
cerebral edema owing to acute hyponatremia(2) and osmotic
demyelination owing to overly rapid correction of chronic
hyponatremia.(3) However, mild chronic hyponatremia traditionally has been regarded as benign.(4) Recent studies suggest
a need for reappraisal of this view. For example, Renneboog
and colleagues showed that patients in whom hyponatremia
had been classified as chronic and asymptomatic actually had
various neurologic deficits when analyzed more closely.(5) In this
matched case-control study, the authors showed that patients
with chronic hyponatremia fell four times more often than
controls, even though acute illnesses, chronic conditions,
polypharmacy, use of central nervous system drugs, and

vasodilators were not more common.(5) A follow-up study by

the same group demonstrated that mild hyponatremia (serum
sodium concentration 131
3 mmol/L) was more prevalent in
ambulatory elderly subjects with bone fractures after incidental
falls.(6) Similarly, using a retrospective cohort, Kinsella and
colleagues found hyponatremia to be associated with fracture
occurrence independent of osteoporotic risk factors and
osteoporosis treatment.(7) Interestingly, Verbalis and colleagues
added yet another perspective by showing that hyponatremia by
itself is a risk factor for osteoporosis (low bone mineral density).(8)
This conclusion was drawn from a cross-sectional analysis of
the Third National Health and Nutrition Examination Survey and
was further supported by demonstrating the development of
osteoporosis in an animal model of chronic hyponatremia.(8)
The relationship between mild hyponatremia and the incidence
of fractures has not been studied prospectively in a populationbased study. In summary, at present, it is still unknown if
hyponatremia is associated with fractures and, if so, whether this

Received in original form August 29, 2010; revised form December 7, 2010; accepted February 18, 2011. Published online March 4, 2011.
Address correspondence to: M Carola Zillikens, MD, PhD, PO Box 2040, Room D-428, 3000 CA Rotterdam, The Netherlands. E-mail: m.c.zillikens@erasmusmc.nl
Journal of Bone and Mineral Research, Vol. 26, No. 8, August 2011, pp 18221828
DOI: 10.1002/jbmr.380
2011 American Society for Bone and Mineral Research

1822

is because of falls, osteoporosis, or both. Therefore, we analyzed

in the Rotterdam Study, a prospective, population-based cohort
study,(9) the possible relationships between hyponatremia, bone
mineral density (BMD), falls, and fractures. Since hyponatremia
has been associated with increased mortality, which may lead to
underestimation of fracture risk, we also analyzed the relationship between hyponatremia and all-cause mortality.(10)

getting in and out of a car. The index is represented by a

continuous score ranging from 0 to 3, where 0 indicates no
impairment and 3 indicates severe impairment.(15) The use of
medication, including drugs potentially associated with hyponatremia, diuretics, and psycholeptics,(16) was determined at
baseline by interview and verified by a physician.

Fractures

Materials and Methods

Study population
This study was conducted within the framework of the
Rotterdam Study, an ongoing prospective, population-based
cohort study among subjects aged 55 years and older, living in
Ommoord, a suburb of Rotterdam, The Netherlands. The
rationale and design of the Rotterdam Study have been
described elsewhere.(9,11,12) All 10,275 inhabitants were invited
for baseline examination between August 1990 and June 1993.
Of those, 7983 participated. For this study, we analyzed a total of
5208 participants for whom data on serum sodium concentration were present at baseline. The study was approved by the
Medical Ethics Committee of the Erasmus Medical Center, and
written informed consent was obtained from all participants.

Measurements
Height (cm) and weight (kg) were measured at the initial
examination in standing position wearing indoor clothes without
shoes. Body mass index (BMI) was calculated as weight in
kilograms divided by height in centimeters squared (kg/cm2).
BMD (in g/cm2) of the hip and lumbar spine (L2L4) was
measured by dual-energy X-ray absorptiometry (DXA) using a
Lunar DPX densitometry apparatus (Lunar Corp., Madison, WI,
USA). Methods, quality assurance, and accuracy of the DXA
measurements have been described previously.(13) Serum
sodium and nonfasting glucose levels were measured within
our clinical chemistry department using standard methods.
Hyponatremia was defined as a serum sodium concentration of
less than 136 mmol/L, which is the cutoff used in the laboratory
in which all samples were measured.

Potential confounders
The presence of type 2 diabetes mellitus was defined by the
current use of antidiabetic medication or by a nonfasting or
postload plasma glucose level concentration above 200 mg/dL
(11.1 mmol/L). Prevalence of myocardial infarction was defined
according to the International Classification of Diseases, 10th
revision (ICD-10).(14) Dementia was diagnosed with the use of the
MiniMental State Examination and the Geriatric Mental State
Schedule.(14) The number of falls in the preceding year and
current smoking status were assessed with the use of a
questionnaire. Falling also was assessed using structured
personal interviews by trained medical research nurses. A faller
was defined as an individual with a history of one or more falls
without precipitating trauma (eg, car accident or sports injury) in
the 12 months preceding the baseline interview. A lower limb
disability index was obtained by calculating the mean score of
answers to questions concerning rising, walking, bending, and
HYPONATREMIA AND RISK OF FRACTURES

Fracture events were obtained from the computerized records of

the general practitioners in the research area. Research
physicians regularly followed participant information in the
general practitioners records outside the research area and
made an independent review and encoding of all reported
events. Subsequently, a medical expert reviewed all coded
events for the final classification using ICD-10.(14) Additional
information on hip fractures was gathered through the Dutch
National Hospital Registration System. An osteoporosis expert
reviewed all coded events for final classification. Incident
fractures were analyzed with and without excluding fractures
caused by cancer and all hand, foot, skull, and face fractures.
Information on incident nonvertebral fractures has been
collected within an average follow-up period of 7.4
3.3 years.
The period of follow-up was calculated as the time from
enrollment in the study to the first fracture, death, or the end of
the planned follow-up period, whichever occurred first. For
subjects lost to the study during follow-up, the follow-up period
was censored at the time of the last contact with the subject. For
this study, follow-up ended on January 1, 2002. Procedures for
all-cause mortality follow-up were described previously, and this
analysis also was based on data until January 1, 2002.(17)

Vertebral fracture assessment

Both at baseline (19901993) and at the second follow-up visit
(between 1997 and 1999), radiographs of the thoracolumbar
spine were available for 2526 individuals with serum sodium
determinations at baseline with a mean follow-up of 6.4 years. All
thoracolumbar spine radiographs of the follow-up visit were
scored for the presence of vertebral fracture using the
McCloskey/Kanis method, as described previously.(18) If vertebral
fractures were detected, the baseline radiograph also was
evaluated. If the vertebral fracture was already present at
baseline, it was considered to be a prevalent fracture. If it was not
present at baseline, the fracture was defined as incident.

Statistical analysis
Analysis of variance (ANOVA) was used to examine the
associations between baseline general characteristics across
subjects with normonatremia and hyponatremia. Analysis of
covariance (ANCOVA) was performed to adjust for age and sex.
Comparison of BMD was made with and without factors that can
cause hyponatremia, including type 2 diabetes mellitus and
diuretics.(1921) Coxs proportional hazards regression was used
to estimate the relative risk of nonvertebral fractures and also of
all-cause mortality. Proportionality of hazards was checked.
Logistic regression was used to estimate the risk of vertebral
fractures and recent falls at baseline. All analyses were adjusted
by the a priori defined factors age, sex, and BMI. We also
Journal of Bone and Mineral Research

1823

performed a fully adjusted multivariate analysis using the

following covariates: sex, age, BMI, disability index, use of
diuretics, use of psycholeptics, prevalent diabetes, and recent
falls (except when the outcome was recent falls). We used
missing-value indicators for missing values of cofactors. All
analyses were done using SPSS (Version 15.0; SPSS, Inc., Chicago,
IL, USA). p Values lower than .05 were considered significant. All
data are expressed as mean
SD.

Results
General characteristics
The general characteristics of the study population are presented
in Table 1. The majority of the population (61.5%) was female,
and the mean age was 70.3
9.1 years. Hyponatremia was
detected in 399 subjects (7.7%, serum sodium concentration
133.4
2.0 mmol/L, lowest level 124 mmol/L). The distribution of
serum sodium in the study population is shown in Fig. 1.

Baseline comparison of subjects with and without

hyponatremia
Table 2 compares the general characteristics, comorbidity, and
BMD at baseline in subjects with normonatremia with those with
hyponatremia. Subjects with hyponatremia were significantly
older (73.5
10.3 years versus 70.0
9.0 years, p < .001), had a
2 times higher prevalence of type 2 diabetes mellitus (22.2%
versus 10.3%, p < .001), and used diuretics more often (31.1%
versus 15.0%, p < .001). Nonfasting serum glucose concentration
was significantly higher in subjects with hyponatremia
[159
95 mg/dL versus 124
43 mg/dL (8.8
5.3 mmol/L versus
6.9
2.4 mmol/L), p < .001]. All types of diuretics were used
significantly more often in patients with hyponatremia (thiazide
diuretics 18% versus 10%, loop diuretics 11% versus 5%,
potassium-sparing diuretics 6% versus 1%, all p < .001). The
higher prevalence of dementia, myocardial infarction, and
peripheral artery disease in subjects with hyponatremia at
baseline disappeared after adjustment for age and sex.
Conversely, disability index, systolic and diastolic blood pressure,
and height remained slightly but significantly higher in subjects
Table 1. Baseline Characteristics of the Study Population
Total population
(n 5208)
Women, n (%)
Age, years
Height (cm) (n 5042)a
Weight (kg) (n 5057)a
Body mass index (kg/m2) (n 5038)a
Current smoking, n (%)
Use of diuretics, n (%)
Serum sodium, mmol/L (range)
Serum sodium < 136 mmol/L
Femoral neck BMD (g/cm2) (n 4185)a
Lumbar spine BMD (g/cm2) (n 4195)a

3204 (61.5)
70.27
9.13
166.14
9.28
72.53
11.92
26.27
3.78
1199 (23.0)
861 (16.5)
140.2
3.3 (124160)
399 (7.7)
0.86
0.14
1.09
0.20

BMD bone mineral density.

a
The number of subjects in whom these data were available.

1824

Journal of Bone and Mineral Research

Fig. 1. Distribution of serum sodium in the study population. The

numbers of subjects that had a serum sodium in the lowest or highest
ranges were as follows: n 1 for serum sodium 124, 125, 155, 156, 157,
158, 159, and 160 mmol/L; n 2 for serum sodium 126 and 154 mmol/L;
and n 3 for serum sodium 127 mmol/L. Owing to the scaling, these
subjects are not visible in the graph.

with hyponatremia at baseline after adjustment for age and sex.

Subjects with hyponatremia had more recent falls at baseline
(23.8% versus 16.4%, p < .01). Multivariate logistic regression
showed that the increased risk of falls for hyponatremic subjects
remained after adjusting for age, sex, BMI, disability index, use of
diuretics or psycholeptics, and prevalent diabetes mellitus [odds
ratio (OR) 1.35, 95% confidence interval (CI) 1.031.75,
p .029]. Hyponatremia was not associated with lower femoral
neck or lumbar spine BMD. In fact, femoral neck BMD was slightly
but not significantly higher in subjects with hyponatremia
(0.875
0.15 g/cm2) than in subjects with normonatremia
(0.861
0.14 g/cm2, p 0.1) after adjustment for age and sex.
After exclusion of subjects with type 2 diabetes mellitus
and subjects using diuretics, results were not substantially
altered (femoral neck BMD in subjects with hyponatremia
0.877
0.14 g/cm2 and in subjects with normonatremia
0.855
0.14 g/cm2, p .09); similar results were obtained when
only subjects with hyponatremia and thiazide use were excluded
(data not shown). Finally, we performed a linear regression to
analyze whether serum sodium concentration (as a continuous
parameter) predicted BMD in the whole population, in subjects
with normonatremia, or in subjects with hyponatremia; this
analysis also showed no relationship between serum sodium
concentration and BMD (data not shown).

Fracture risk
Table 3 presents the association of hyponatremia with fractures.
Hyponatremia was related to increased risk of incident
nonvertebral fractures [hazard ratio (HR) 1.39, 95% CI 1.11
1.73, p .004] after adjustment for age, sex, and BMI. Further
adjustments for disability index, use of diuretics or psycholeptics,
recent falls, and prevalent diabetes did not modify the results
(HR 1.34, 95% CI 1.081.68, p .009). The results also did not
HOORN ET AL.

Table 2. Comparison of Baseline Characteristics, Comorbidity, Medication, and BMD of Subjects With Normal and Decreased Serum
Sodium (<136 mmol/L)

Characteristic
General characteristics
Women, n (%)
Age (years)
Height (cm)
Weight (kg)
Body mass index (kg/m2)
Systolic blood pressure (mm Hg)
Diastolic blood pressure (mm Hg)
Comorbidity and medication
Current smoking, n/total (%)
Diabetes mellitus at baseline, n/total (%)
Dementia at baseline, n/total (%)
MI at baseline, n/total (%)
Peripheral artery disease, n/total (%)
Recent fall, n/total (%)
Disability index
Disability index  0.5, n/total (%)
Use of diuretics, n/total (%)
Use of psycholeptics, n/total (%)
BMD (g/cm2)
Femoral neck BMD
Lumbar spine BMD

Normonatremia
(n 4809)

Hyponatremia
(n 399)

2952 (61.4)
70.0
9.0
166.3
9.2
72.7
11.9
26.3
3.8
139.6
22.2
73.7
11.8

252 (63.2)
73.5
10.3
164.6
10.0
70.9
12.4
26.1
3.7
138.7
23.2
71.9
12.4

1116/4671 (23.9)
491/4789 (10.3)
245/4571 (5.4)
609/4486 (13.6)
854/4286 (19.9)
787 (16.4)
0.62 (1.39)
1557 (32.4)
521/3471 (15.0)
108/3471 (3.1)

83/352 (23.6)
86/388 (22.2)
32/363 (8.8)
54/311 (17.4)
106/370 (28.6)
95 (23.8)
0.97 (1.76)
179 (44.9)
103/331 (31.1)
8/331 (2.4)

0.861
0.14
1.087
0.20

0.873
0.15
1.083
0.21

p Values
(ANOVA)
.997
<.001
.001
.006
.452
.477
.003
.894
<.0001
.0006
.024
<.0001
.00013
.000003
<.0001
.0003
.502
.205
.770

p Values
(ANCOVA)a

.025
.103
.426
.014
.038
.353
<.0001
.228
.122
.056
.006
.031
.037
<.0001
.483
.105
.473

BMD bone mineral density; MI myocardial infarction.

a
Adjusted for age and sex.

Table 3. Association of Hyponatremia With Fractures

Fractures
Nonvertebral
Normonatremia
Hyponatremia
Hip
Normonatremia
Hyponatremia
Vertebral totalb
Normonatremia
Hyponatremia
Vertebral prevalent
Normonatremia
Hyponatremia
Vertebral incident
Normonatremia
Hyponatremia

Modela

1
2

1
2

1
2

1
2

1
2

No. of fractures, total (%)

833/4809 (17.3)
93/399 (23.3)

214/4809 (4.4)
31/399 (7.8)

269/2390 (11.3)
23/136 (16.9)

177/2390 (7.4)
17/136 (12.5)

131/2390 (5.5)
8/136 (5.9)

Relative risk (95% CI)

HR
Reference
1.39 (1.111.73)
1.34 (1.081.68)
HR
Reference
1.44 (0.972.15)
1.39 (0.932.08)
OR
Reference
1.54 (0.962.45)
1.61 (1.002.59)
OR
Reference
1.69 (0.992.89)
1.78 (1.043.06)
OR
Reference
1.03 (0.492.16)
1.08 (0.512.27)

p Value

.004
.009

.071
.105

.073
.049

.057
.037

.94
.85

a
Model 1: Adjusted for age, sex, and body mass index. Model 2: Adjusted for age, sex, body mass index, disability index, use of diuretics, use of
psycholeptics, prevalent diabetes mellitus, and recent falls.
b
Vertebral total equals vertebral prevalent and incident. Numbers do not sum up because subjects may have both prevalent and incident fractures.
Abbreviations: CI, confidence interval; HR, hazard ratio; No., number; OR, odds ratio.

HYPONATREMIA AND RISK OF FRACTURES

Journal of Bone and Mineral Research

1825

change after excluding fractures caused by cancer and hand,

foot, skull, and face fractures (HR 1.35, 95% CI 1.081.70,
p .009). When hip fractures were considered separately, the
presence of hyponatremia was associated with a borderline
significant increased risk after adjustment for age, sex, and BMI
(HR 1.44, 95% CI 0.972.15, p .07). Subjects with hyponatremia also had an increased risk of vertebral fractures
(combined prevalent and incident) after adjustment for all
covariates (OR 1.61, 95% CI 1.002.59, p .049). There was
no significant association with incident vertebral fractures.
Additional adjustment for BMD or the different types of diuretics
did not modify any of the associations with fractures (data not
shown).

Hyponatremia and all-cause mortality

All-cause mortality was higher in subjects with hyponatremia
(206 of 399, or 51.6%, versus 1567 of 4809, or 32.6%, p < .001).
An increased risk of all-cause mortality in subjects with
hyponatremia remained after adjustment for age, sex, and
BMI (HR 1.35, 95% CI 1.151.57, p .00017). After further
adjustment for comorbidity at baseline (ie, diabetes, dementia,
myocardial infarction, hypertension, falling, and use of diuretics),
the HR of all-cause mortality was 1.21 (95% CI 1.031.43,
p .022). The increase in (unadjusted) mortality risk is shown in
the Kaplan-Meier curve of all-cause mortality in subjects with
mild hyponatremia (dark gray) compared with those with normal
serum sodium concentrations (light gray) (Fig. 2).

Discussion
Recent studies suggest that mild hyponatremia is associated
with fractures owing to increased risk of falling or a decreased
BMD.(58) However, these studies were retrospective,(7) case-

Fig. 2. Kaplan-Meier curve showing mortality at follow-up in subjects

with hyponatremia and normonatremia. The curves of subjects with
normonatremia and hyponatremia (serum sodium < 136 mmol/L) are
both indicated and shown in light gray and darker gray, respectively.

1826

Journal of Bone and Mineral Research

controlled,(4,6) or cross-sectional,(8) but not prospective. Therefore, we conducted a population-based prospective cohort study
in the elderly that investigated whether hyponatremia was
associated with fractures and, if so, whether this effect was
mediated by decreased BMD and/or falls. Our principal results
indeed indicate that subjects with mild hyponatremia had an
increased risk of incident nonvertebral fractures and prevalent
but not incident vertebral fractures. Hyponatremia was not
related to femoral neck or lumbar spine BMD (Table 2), and we
were therefore unable to replicate an association between
hyponatremia and osteoporosis.(8) It is possible that the
development of osteoporosis (low BMD) requires a more
severe degree of hyponatremia, as suggested by the study by
Verbalis and colleagues.(8) Indeed, in our population, 93% of
hyponatremic subjects had a serum sodium concentration
between 130 and 135 mmol/L (Fig. 1). The association between
hyponatremia and fracture risk in our study was independent of
disability index, use of diuretics or psycholeptics, recent falls, and
diabetes mellitus.
The focus of this study was on the complications (ie, fractures,
mortality) rather than on the causes of hyponatremia.
Nevertheless, the possible etiology of hyponatremia merits
consideration because it remains unclear whether hyponatremia
by itself predisposes to fractures or whether hyponatremia is a
surrogate marker for another risk factor for fractures. The higher
frequency of use of diuretics in subjects with hyponatremia
suggests that diuretics were a common cause of hyponatremia.
Indeed, thiazides and, to a lesser extent, potassium-sparing
diuretics are a common cause of hyponatremia.(22) The use of
psycholeptics at baseline, another common cause of hyponatremia,(16) was low and not higher in subjects with hyponatremia.
Diabetes mellitus is another well-known cause of hyponatremia
either because of hyperglycemia(23) or because of altered
vasopressin metabolism.(24) Serum glucose concentration was
significantly higher in patients with hyponatremia, but the
degree of hyperglycemia could account for only a 18 to 36 mg/dL
(1 to 2 mmol/L) fall in serum sodium using a generally accepted
correction factor.(23)
Both diuretics and diabetes mellitus can affect BMD and
fracture risk. Thiazide diuretics have been shown to reduce the
risk of hip fractures(25) and to increase BMD,(21) possibly owing to
a direct effect on bone.(26) Conversely, loop diuretics have been
associated with hip bone loss in older men(20) and increased risk
of fractures in postmenopausal women.(27) Subjects with type 2
diabetes mellitus have a higher fracture risk despite a higher
BMD.(19,28) Therefore, in theory, thiazide diuretics and diabetes
mellitus could have counteracted a possible lowering of BMD by
hyponatremia. However, what argues against this possibility is
that BMD was not lower in subjects with hyponatremia after
excluding subjects with type 2 diabetes mellitus and subjects
using diuretics in general or thiazide diuretics in particular. Liver
cirrhosis and congestive heart failure also can cause hyponatremia and also have been implicated in osteoporosis,(29,30) but
this most likely would have affected BMD. The syndrome of
inappropriate antidiuresis, another common cause of hyponatremia, can be paraneoplastic,(31) but the types of fractures we
observed were not typical for pathologic fractures, and results
were not modified after exclusion of fractures owing to cancer. In
HOORN ET AL.

summary, no single underlying factor associated with hyponatremia emerged, suggesting that hyponatremia by itself should
be considered a risk factor for fractures.
This conclusion readdresses the pivotal question of what
the possible explanations may be for the association between
hyponatremia and fractures. Interestingly, this is the first study to
confirm the recently identified relationship between falls and
hyponatremia.(5) Because nonvertebral fractures usually require
a fall, the association between hyponatremia and the risk of
nonvertebral fractures does suggest a role for falls. However,
adjustment for recent falls at baseline did not modify the results,
suggesting that falling does not explain the association. Also,
the association between hyponatremia and vertebral fractures
argues against a primary role for falling because vertebral
fractures are usually not traumatic. Therefore, alternatively or in
addition to falls, an unknown factor influencing bone quality but
not BMD also could account for our findings. For example, it is
known that damaged bone generates a steady electric (ionic)
current at the damaged site and that sodium is one of the
carriers of this current.(32) One thus might speculate that
hyponatremia indirectly results in decreased local repair of
microdamage to the bone, resulting in increased fracture risk
without an effect on BMD. Alternatively, one could hypothesize
that lower serum osmolality resulting from hyoponatremia
influences collagen composition of bone. In articular cartilage,
osmotic pressure has been shown to influences collagen fiber
architecture.(33)
Our data add to the growing body of evidence indicating that
chronic hyponatremia is not benign.(48,10,34) Although the vast
majority of studies on the complications of hyponatremia were
conducted in hospitalized patients, a recent study showed that
hyponatremia also carries a poor prognosis in communitydwelling subjects, including an increased risk of myocardial
infarction and death.(34) Here, we also showed that hyponatremia was associated with higher all-cause mortality (Fig. 2). This
may even have led to an underestimation of the effect of
hyponatremia on fracture risk because subjects with hyponatremia may have died before a fracture could have occurred. An
unresolved question is whether hyponatremia is the cause of
increased mortality or whether hyponatremia is reflecting the
severity of underlying disorders that increased mortality. For
example, hyponatremia in liver cirrhosis(35) and congestive heart
failure(36) has long been recognized as a marker for poor
prognosis but appears to reflect the severity of these disorders.
However, recently, a large single-center study showed that
both community- and hospital-acquired hyponatremia were
associated with increased mortality, even when correcting for
comorbidity and medication.(10) Similarly, in our study, the
association between hyponatremia and mortality did not
disappear after adjustment for comorbidity. Therefore, chronic
mild hyponatremia no longer should be regarded as a benign
disorder.
An important limitation of our study is that we only measured
serum sodium concentration at baseline and therefore do not
know whether hyponatremia was transient or chronic. However,
the fact that already a single hyponatremic value revealed an
association with fracture risk suggests that the association is
quite strong. Before the association between hyponatremia and
HYPONATREMIA AND RISK OF FRACTURES

fractures can be generalized, our results should be replicated in

other prospective studies.
In conclusion, we observed that elderly subjects with mild
hyponatremia at baseline had an increased risk of incident
nonvertebral fractures and of prevalent but not incident
vertebral fractures. The increased fracture risk was not explained
by increased comorbidity, the use of diuretics, or a history of
recent falling. Mild hyponatremia is a new and important risk
factor for fractures in the elderly. When our findings are
replicated in an independent population, screening for hyponatremia should be advised for elderly patients with fractures.
Furthermore, our findings suggest that mild hyponatremia in the
elderly should be treated to prevent complications such as
fractures.

Disclosures
All the authors state that they have no conflicts of interest.

Acknowledgments
The Rotterdam Study is funded by the Erasmus Medical Center
and Erasmus University; the Rotterdam, Netherlands Organization for the Health Research and Development (ZonMw); the
Research Institute for Diseases in the Elderly (RIDE); the Ministry
of Education, Culture and Science; the Ministry for Health,
Welfare and Sports; the European Commission (DG XII); and
the Municipality of Rotterdam. The authors are very grateful
to the participants and staff from the Rotterdam Study and the
participating general practitioners and pharmacists.

References
1. Adrogue HJ, Madias NE. Hyponatremia. N Engl J Med. 2000;342:1581
1589.
2. Arieff AI. Hyponatremia, convulsions, respiratory arrest, and permanent brain damage after elective surgery in healthy women. N Engl J
Med. 1986;314:15291535.
3. Sterns RH, Riggs JE, Schochet SS Jr. Osmotic demyelination syndrome
following correction of hyponatremia. N Engl J Med. 1986;314:1535
1542.
4. Decaux G. Is asymptomatic hyponatremia really asymptomatic? Am J
Med. 2006;119:S7982.
5. Renneboog B, Musch W, Vandemergel X, Manto MU, Decaux G. Mild
chronic hyponatremia is associated with falls, unsteadiness, and
attention deficits. Am J Med. 2006;119:71e18.
6. Gankam Kengne F, Andres C, Sattar L, Melot C, Decaux G. Mild
hyponatremia and risk of fracture in the ambulatory elderly. QJM.
2008;101:583588.
7. Kinsella S, Moran S, Sullivan MO, Molloy MG, Eustace JA. Hyponatremia independent of osteoporosis is associated with fracture
occurrence. Clin J Am Soc Nephrol. 2010;5:275280.
8. Verbalis JG, Barsony J, Sugimura Y, Tian Y, Adams DJ, Carter EA, et al.
Hyponatremia-Induced Osteoporosis. J Bone Miner Res. 2010;25:
554563.
9. Hofman A, Breteler MM, van Duijn CM, et al. The Rotterdam Study:
2010 objectives and design update. Eur J Epidemiol. 2009;24:553
572.

Journal of Bone and Mineral Research

1827

10. Wald R, Jaber BL, Price LL, Upadhyay A, Madias NE. Impact of hospitalassociated hyponatremia on selected outcomes. Arch Intern Med.
2010;170:294302.
11. Hofman A, Breteler MM, van Duijn CM, et al. The Rotterdam Study:
objectives and design update. Eur J Epidemiol. 2007;22:819829.
12. Hofman A, Grobbee DE, de Jong PT, van den Ouweland FA. Determinants of disease and disability in the elderly: the Rotterdam Elderly
Study. Eur J Epidemiol. 1991;7:403422.
13. Burger H, van Daele PL, Algra D, et al. The association between age
and bone mineral density in men and women aged 55 years and
over: the Rotterdam Study. Bone Miner. 1994;25:113.
14. WHO. International Statistical Classification of Diseases and Related
Health Problems. WHO. 10th ed. Geneva, Switzerland; 1992.
15. Odding E, Valkenburg HA, Grobbee DE, Hofman A, Pols HA. [Locomotor disability in the elderly; the ERGO Study (Erasmus Rotterdam
Health and the Elderly). ERGO Study Group]. Ned Tijdschr Geneeskd.
1995;139:20962100.
16. Liamis G, Milionis H, Elisaf M. A review of drug-induced hyponatremia. Am J Kidney Dis. 2008;52:144153.
17. Sayed-Tabatabaei FA, Schut AF, Vasquez AA, et al. Angiotensin
converting enzyme gene polymorphism and cardiovascular
morbidity and mortality: the Rotterdam Study. J Med Genet. 2005;
42:2630.
18. McCloskey EV, Spector TD, Eyres KS, et al. The assessment of vertebral
deformity: a method for use in population studies and clinical trials.
Osteoporos Int. 1993;3:138147.
19. de Liefde I, van der Klift M, de Laet CE, van Daele PL, Hofman A, Pols
HA. Bone mineral density and fracture risk in type-2 diabetes mellitus:
the Rotterdam Study. Osteoporos Int. 2005;16:17131720.
20. Lim LS, Fink HA, Kuskowski MA, Taylor BC, Schousboe JT, Ensrud KE.
Loop diuretic use and increased rates of hip bone loss in older men:
the Osteoporotic Fractures in Men Study. Arch Intern Med. 2008;
168:735740.

23. Matz R. Hyponatremia: evaluating the correction factor for hyperglycemia. Am J Med. 2000;108:180181.
24. Bankir L, Bardoux P, Ahloulay M. Vasopressin and diabetes mellitus.
Nephron. 2000;87:818.
25. Schoofs MW, van der Klift M, Hofman A, de Laet CE, Herings RM,
Stijnen T, et al. Thiazide diuretics and the risk for hip fracture. Ann
Intern Med. 2003;139:476482.
26. Dvorak MM, De Joussineau C, Carter DH, et al. Thiazide diuretics
directly induce osteoblast differentiation and mineralized nodule
formation by interacting with a sodium chloride co-transporter in
bone. J Am Soc Nephrol. 2007;18:25092516.
27. Carbone LD, Johnson KC, Bush AJ, et al. Loop diuretic use and fracture
in postmenopausal women: findings from the Womens Health
Initiative. Arch Intern Med. 2009;169:132140.
28. Yamamoto M, Yamaguchi T, Yamauchi M, Kaji H, Sugimoto T. Diabetic
patients have an increased risk of vertebral fractures independent of
BMD or diabetic complications. J Bone Miner Res. 2009;24:702709.
29. Abou-Raya S, Abou-Raya A. Osteoporosis and congestive heart failure
(CHF) in the elderly patient: double disease burden. Arch Gerontol
Geriatr. 2009;49:250254.
30. Giouleme OI, Vyzantiadis TA, Nikolaidis NL, et al. Pathogenesis of
osteoporosis in liver cirrhosis. Hepato-gastroenterology. 2006;53:
938943.
31. Verbalis JG. Tumoral hyponatremia. Arch Intern Med. 1986;146:1686
1687.
32. Borgens RB. Endogenous ionic currents traverse intact and damaged
bone. Science. 1984;225:478482.
33. Korhonen RK, Han SK, Herzog W. Osmotic loading of articular cartilage modulates cell deformations along primary collagen fibril directions. J Biomech. 2010;43:783787.
34. Sajadieh A, Binici Z, Mouridsen MR, Nielsen OW, Hansen JF, Haugaard
SB. Mild hyponatremia carries a poor prognosis in community
subjects. Am J Med. 2009;122:679686.

21. Morton DJ, Barrett-Connor EL, Edelstein SL. Thiazides and bone
mineral density in elderly men and women. Am J Epidemiol. 1994;
139:11071115.

35. Kim WR, Biggins SW, Kremers WK, et al. Hyponatremia and mortality
among patients on the liver-transplant waiting list. N Engl J Med.
2008;359:10181026.

22. Friedman E, Shadel M, Halkin H, Farfel Z. Thiazide-induced hyponatremia. Reproducibility by single dose rechallenge and an analysis of
pathogenesis. Ann Intern Med. 1989;110:2430.

36. Gheorghiade M, Rossi JS, Cotts W, et al. Characterization and prognostic value of persistent hyponatremia in patients with severe heart
failure in the ESCAPE Trial. Arch Intern Med. 2007;167:19982005.

1828

Journal of Bone and Mineral Research

HOORN ET AL.