Fax +41 61 306 12 34

E-Mail [email protected]
www.karger.com
In-Depth Topic Review
Am J Nephrol 2012;36:97104
DOI: 10.1159/000339625
Understanding Hypernatremia
Ramin Sam
a
Iraj Feizi
b


a
Division of Nephrology, San Francisco General Hospital and the University of California, San Francisco, Calif. , USA;

b
Division of Surgery, Ardabil University of Medical Sciences, Ardabil , Iran

pernatremia not only how much fluid to give is ques-
tioned but also what kind of fluid. With hyponatremia,
the problem is usually figuring out what has led to the
abnormal serum sodium concentration. In order to undo
some of this confusion, one first needs to understand why
sodium concentration problems are so difficult.
Sodium concentration problems are difficult because
as far as sodium is concerned the body is comprised of at
least two compartments (the extracellular and the intra-
cellular fluid). If one imagines a one-compartment sys-
tem where sodium concentration is same in the total
body water, sodium concentration problems would be
simple ( fig.1 ). In a one-compartment system, the serum
sodium concentration is equal to total body sodium di-
vided by total body water.
[Na] = total body Na/total body water (in a one-compart-
ment system)
Imagining this one-compartment system, if the serum
sodium concentration is for example 160 meq/L and total
body water is 40 liter, one can easily see that administer-
ing 5.7 liter of pure water will lower the serum sodium
concentration to 140 meq/L (assuming that there is no
sodium or water loss, and no sodium intake). Again, in a
one-compartment system if one starts out with a serum
sodium concentration of 140 meq/L (total body water 40
liter), makes 2 liter of urine with sodium concentration of
40 meq/L in 24 h, and has no intake, then the serum so-
Key Words
Hypernatremia Hyponatremia Sodium and water
balance
Abstract
Understanding hypernatremia is at times difficult for many
clinicians. However, hypernatremia can often be deciphered
easily with some basic understanding of water and sodium
balance. Here, the basic pathophysiological abnormalities
underlying the development of sodium disorders are re-
viewed, and case examples are given. Hypernatremia often
arises in the hospital, especially in the intensive care units
due to the combination of (1) not being able to drink water;
(2) inability to concentrate the urine (most often from having
kidney failure); (3) osmotic diuresis from having high serum
urea concentrations, and (4) large urine or stool outputs.
Copyright 2012 S. Karger AG, Basel
Why Are Serum Sodium Concentration Problems
Difficult?
Hyponatremia and hypernatremia are commonly en-
countered problems both in the outpatient setting and
especially in hospitalized patients. Frequently, confusion
arises as to how best manage these patients with serum
sodium concentration abnormalities. In the case of hy-
Received: May 16, 2012
Accepted: May 21, 2012
Published online: June 27, 2012
Nephrology
American Journal of
Ramin Sam, MD
San Francisco General Hospital
1001 Potrero Ave, Building 100, Rm 349
San Francisco, CA 94110-1341 (USA)
Tel. +1 415 206 6605, E-Mail samr @ medsfgh.ucsf.edu
2012 S. Karger AG, Basel
02508095/12/03610097$38.00/0
Accessible online at:
www.karger.com/ajn
Sam /Feizi Am J Nephrol 2012;36:97104 98
dium concentration at the end of the 24-hour period
would be 145 meq/L. In summary, in a one-compartment
system, all sodium concentration problems can be solved
by knowing few simple facts (serum sodium concentra-
tion, total body water, urine volume, urine sodium con-
centration, sodium intake and water intake).
How to Make a One-Compartment System Out of
Two Compartments
However, body fluid is not a one-compartment sys-
tem as far as sodium is concerned, but it is composed of
at least two compartments ( fig.2 ). In 1958, Edelman et
al. [1] came up with a brilliant idea of making a one-
compartment system out of two compartments. Since in-
tracellular potassium concentration is almost the same
as extracellular sodium concentration and vice versa,
they postulated that serum sodium concentration is pro-
portional to total body sodium plus potassium divided
by total body water ([Na] = (total body Na + total body
K)/total body water). They tested their idea and found
good results (correlation coefficient of r = 0.83). Put an-
other way, when we take in potassium which goes most-
ly intracellularly, the serum sodium concentration will
increase by movement of water (and not of sodium) from
extracellular space to intracellular space [13] . As an ex-
ample, if one has a baseline serum sodium concentration
of 140 meq/L and has total body water of 40 liter, then the
total body sodium plus potassium will be 5,600 meq. If
in the next 24 h that person makes 2 liter of urine which
has a sodium concentration of 40 meq/L and a potassium
concentration of 40 meq/L also and has no intake, then
the serum sodium concentration will be 143 meq/L at the
end of the 24-hour period. Understanding that this for-
mula is not exact because of a variety of reasons is im-
portant. In this system, one will need the total body wa-
ter, intake of potassium, sodium, and water, output of
potassium, sodium and water and beginning serum so-
dium concentration to solve the sodium concentration
problem.
The Other Compartments
Because of the Gibbs-Donnan effect, and the fact that
electro-neutrality has to be maintained both in the intra-
vascular and in the interstitial fluid, the interstitial fluid
has a sodium concentration slightly less than the intra-
vascular fluid (139 meq/L rather than 142 meq/L) [4] . The
Gibbs-Donnan effect arises from the fact that the serum
albumin concentration is higher in the intravascular flu-
id than the interstitial fluid. Thus, in order to maintain
electroneutrality, as albumin is net negatively charged,
there has to be a little more sodium in the intravascular
space than the interstitial space. Thus, the interstitial flu-
id will constitute a third compartment.
The fourth compartment is the sodium bound to pro-
teins, stored in the skin or in the bone which is in equi-
librium with the rest of the body sodium but is not os-
motically active [5] . This compartment also contributes
to the serum sodium concentration as it has an effect on
the ability of sodium intake to raise serum sodium con-
centration. For instance, if this compartment did not ex-
ist, intake of 100 meq of NaCl may increase serum sodi-
um from 142 to 144 meq/L, but now 120 meq of NaCl is
necessary to cause the same increase in serum sodium
concentration because 20 meq of the NaCl given, has
been stored in the skin. The last compartment is the so-
dium deep inside the bone which is not in equilibrium
with the rest of the body water sodium and is not a deter-
minant of serum sodium concentration.
Na = 140
meq/L
Water
intake
Water
output
Na
intake
Na
output
Fig. 1. Sodium concentration in a one-compartment system.
Na = 14 meq/L
Intracellular space =
55% of total
body water [14]
Na = 140 meq/L
Extracellular space =
45% of total
body water
ECF ICF
Fig. 2. A two-compartment model.
Sodium Disorders Am J Nephrol 2012;36:97104 99
Modifications of Edelmans Equation
The original Edelman equation is: S
Na
= 1.11 (total
body Na + total body K)/total body water 25.6 [1] . This
equation was simplified in 1986 by Rose to S
Na
= (total
exchangeable body Na + total exchangeable body K)/to-
tal body water [6] . The modified formula has become the
preferred formula clinically, and from it the electrolyte-
free water excretion formula is derived. The formula for
electrolyte-free water excretion is V ! {1 (U
Na
+ U
K
)/
[Na]}, where V is the flow rate of urine. This formula
determines if part of the urine has the same sodium plus
potassium concentration as the serum sodium concen-
tration, how much urine is just free water. For example,
if urine output is 2 l/day, urine Na concentration is 35
meq/L, urine potassium concentration is also 35 meq/L,
and serum sodium concentration is 140 meq/L, then
half of the urine has a urine sodium plus potassium con-
centration of 140 meq/L (the same as the serum sodium
concentration) and half of the urine is just free water.
Thus the electrolyte-free water excretion would be 1 l/
day, meaning that the patient needs to drink 1 l/day of
free water to keep the serum sodium concentration the
same.
More recently, Kurtz and Nguyen have tried to make
sense of slope and the y-intercept of the original Edelman
formula, and have derived more complicated formulas
[78] :
( ) ( )
( )
e e
pw
Na K
Na /
TBW
G

l
= 2

l
l

l
l l
l l
l
l
l
osminactive osminactive
ICF ECF pw +
pw
pw
Na + K
-
TBW
osmol + osmol osmol
+ K +
TBW V
G /

where

pw ISF
pw ISF
V V
V 0 95 V
G
.

q

pw = plasma water, Na
e
= exchangeable Na, K
e
= ex-
changeable potassium, G = Gibbs-Donnan effect, TBW =
total body water, = osmotic coefficient of Na
+
salts,
Na
osm

inactive
= osmotically inactive sodium, K
osm inactive
=
osmotically inactive potassium, osmol
ICF
= osmotically
active non-sodium and non- potassium intracellular os-
moles, osmol
ECF
= osmotically active non-sodium, non-
potassium extracellular osmoles, V
pw
= plasma water vol-
ume, V
ISF
= interstitial fluid volume.
In this equation, G is about 1.04. is for the fact that
sodium salts do not completely dissolve in plasma and
thus the osmolality of Na
+
is not exactly one but about
0.94. G/ is thus 1.11, which is the slope of the original
Edelmans equation.
Most of the equation (the portion in bold) reflects the
y-intercept of 25.6. First, if sodium salt is given, and this
sodium goes inside the osmotically inactive part of the
bone or is osmotically inactive bound to serum proteins,
then it will not raise the serum sodium concentration and
thus needs to be subtracted from the calculated serum
sodium concentration. This sodium does contribute to
the serum sodium concentration though because now we
need more sodium intake to have the same effect on the
serum sodium concentration. Second, if we give a sub-
stance that goes intracellularly or interstitially, that sub-
stance will raise the serum sodium concentration by
drawing water out of the intravascular space and thus
raise the serum sodium concentration (much in the same
way as potassium administration). Third, any substance
(other than sodium) given that stays intravascularly (in-
cluding the small portion of the potassium given) will
tend to lower the serum sodium concentration by draw-
ing water into the intravascular space.
Are Sodium and Potassium Completely Equivalent
in Determining Serum Sodium Concentration?
Edelmans equation assumes sodium and potassium
are completely equivalent in determining serum sodium
concentration. However, this is unlikely to be the case,
although they may be fairly close to being equivalent. In
order to try to answer this question, one has to assume
several factors: (1) a two-compartment system for sodi-
um; (2) ions (sodium) having an osmolality of one (com-
plete dissolution of ions in the plasma); (3) complete os-
motic equilibrium between intravascular, interstitial and
intracellular fluids (may not be complete because of the
Gibbs-Donnan effect); (4) the nonexistence of non-os-
motically active sodium and potassium; (5) no intercom-
partmental movement of sodium or potassium with ad-
ministration of potassium or sodium, respectively, and
(6) intracellular fluid constitutes 55%, extracellular fluid
constitutes 45% of total body water [9] .
With the above assumptions, if one makes a further
assumption that all of the administered KCl will go intra-
cellularly and all of the administered NaCl stays extracel-
lularly, then giving 100 meq of KCl will raise the serum
sodium concentration from 142 to 144.4 meq/L ( fig.3 ). If
one gives 100 meq of KCl (100 mosm K, 100 mosm Cl),
that would raise the plasma osmolality from 300 to 305
mosm/l, as there are now 200 more mosms {[(300 mosm/l
Sam /Feizi Am J Nephrol 2012;36:97104 100
! 40 l) + 200 mosm]/40 l}. If all the KCl goes intracellu-
larly, the total number of osmoles extracellularly has re-
mained the same, and thus the extracellular water has
decreased from 18 to 17.7 l [(300 mosm/l ! 18 l)/305
mosm/l]. The serum sodium concentration has now in-
creased from 142 to 144.4 meq/L [(142 meq/L ! 18 l)/
17.7 l]. Similarly, 100 meq of NaCl will raise it from 142 to
144.3 meq/L ( fig.4 ). Again, giving 100 meq of NaCl will
increase the plasma osmolality from 300 to 305 mosm/l
{[(300 mosm/l ! 40 l) + 200 mosm]/40 l}. If all the NaCl
stays extracellularly, the total number of osmoles intra-
cellularly has remained the same, and thus the intracel-
lular water has decreased from 22 to 21.6 l [(300 mosm/l
! 22 l)/305 mosm/l]. Thus, the extracellular water has
increased from 18 to 18.4 l. The serum sodium concentra-
tion has now increased from 142 to 144.3 meq/L {[(142
meq/L ! 18 l) + 100 meq]/18.4 l}.
If one makes the five assumptions above but assumes
that KCl and NaCl will distribute evenly in the intracel-
lular and extracellular space according to their preexist-
Na = 14
K = 140
Cl = 4
osm = 300
H
2O = 22 l
Na = 142
K = 4.2
Cl = 108
osm = 300
H
2O = 18 l
ECF ICF
100 meq of KCl
Na = 13.8
K = 143
Cl = 8.4
osm = 305
H
2O = 22.3 l
Na = 144.4
K = 4.3
Cl = 110
osm = 305
H
2O = 17.7 l
ECF ICF
Na = 14
K = 140
Cl = 4
osm = 300
H
2O = 22 l
Na = 142
K = 4.2
Cl = 108
osm = 300
H
2O = 18 l
ECF ICF
100 meq of NaCl
Na = 14.3
K = 142.6
Cl = 4.1
osm = 305
H
2O = 21.6 l
Na = 144.3
K = 4.1
Cl = 111
osm = 305
H
2O = 18.4 l
ECF ICF
Fig. 3. Electrolyte concentration before
and after administration of 100 meq of
KCl. Assuming: (1) two-compartment
model, (2) ionic osmolality of 1.0, (3) com-
plete osmotic equilibrium between the
compartments, (4) all administered KCl
goes intracellularly, (5) no intercompart-
mental movement of Na with K adminis-
tration. Ignoring: (1) Gibbs-Donnan ef-
fect, (2) non-osmotically active Na and K.
Fig. 4. Electrolyte concentrations before
and after administration of 100 meq of
NaCl. Assuming: (1) two-compartment
model, (2) ionic osmolality of 1.0, (3) com-
plete osmotic equilibrium between the
compartments, (4) all administered NaCl
goes extracellularly, (5) no intercompart-
mental movement of K with Na adminis-
tration. Ignoring: (1) the Gibbs-Donnan ef-
fect, (2) non-osmotically active Na and K.
3 4
Na = 14
K = 140
Cl = 4
osm = 300
H
2O = 22 l
Na = 142
K = 4.2
Cl = 108
osm = 300
H
2O = 18 l
ECF ICF
100 meq of KCl
98.6 meq intracellular
1.4 meq extracellular
Na = 13.8
K = 142.5
Cl = 8.4
osm = 305
H
2O = 22.3 l
Na = 144.4
K = 4.4
Cl = 109.9
osm = 305
H
2O = 17.7 l
ECF ICF
Na = 14
K = 140
Cl = 4
osm = 300
H
2O = 22 l
Na = 142
K = 4.2
Cl = 108
osm = 300
H
2O = 18 l
ECF ICF
100 meq of NaCl
17 meq intracellular
83 meq extracellular
Na = 14.9
K = 141.3
Cl = 4.8
osm = 305
H
2O = 21.8 l
Na = 145.0
K = 4.2
Cl = 111
osm = 305
H
2O = 18.2 l
ECF ICF
Fig. 5. Electrolyte concentration before and
after administration of 100 meq of KCl. As-
suming: (1) two-compartment model, (2)
ionic osmolality of 1.0, (3) complete osmot-
ic equilibrium between the compartments,
(4) KCl distributes in intracellular and ex-
tracellular fluids, (5) no intercompartmen-
tal movement of Na with K administration.
Ignoring: (1) the Gibbs-Donnan effect, (2)
non-osmotically active Na and K.
Fig. 6. Electrolyte concentrations before and
after administration of 100 meq of NaCl.
Assuming: (1) two-compartment model, (2)
ionic osmolality of 1.0, (3) complete osmotic
equilibrium between the compartments, (4)
NaCl distributes in intracellular and extra-
cellular fluids, (5) no intercompartmental
movement of K with Na administration. Ig-
noring: (1) the Gibbs-Donnan effect, (2)
non-osmotically active Na and K.
5 6
Sodium Disorders Am J Nephrol 2012;36:97104 101
ing concentrations in the compartments, then the follow-
ing will happen. Administration of 100 meq of KCl, will
lead to 98.6 meq to go intracellularly and 1.4 meq to stay
extracellularly. This will lead to an increase in serum so-
dium concentration from 142 to 144.4 too ( fig.5 ). Admin-
istration of 100 meq of NaCl on the other hand, will lead
to 83 meq to stay extracellularly and 17 meq to go intra-
cellularly. The corresponding rise in serum sodium con-
centration would be from 142 to 145 meq/L ( fig.6 ).
Finally, if one assumes the first four assumptions
above, but then assumes that for every 2 molecules of po-
tassium moving intracellularly, 3 molecules of sodium
will move out of the cell as happens with the Na/K ATPase,
then the following will happen. Administration of 100
meq of KCl, will lead to 147.9 meq of Na to move extracel-
lularly, and serum sodium concentration will increase
from 142 to 144.6 meq/L ( fig.7 ).
Although the modifications made to the Edelman
equation and the equivalence of sodium and potassium
help us understand the complexities of sodium concentra-
tion problems, the newer equations are so far not useful
clinically as there are so many assumptions made with
these equations that the simpler formulas may prove to be
just as accurate in predicting serum sodium concentra-
tion.
Normal Sodium Balance
Amazingly, the body is able to maintain serum sodi-
um concentration within a very narrow range (140 8
1 meq/L). Understanding the mechanism of this regula-
tion is crucial in order to comprehend disease states. On
a qualitative basis we take in hyponatric fluid and we ex-
crete hyponatric fluid too, thus maintaining normal se-
rum sodium concentration. Simply, we drink water rath-
er than isotonic saline. This is not always the case with
patients in the intensive care unit unable to drink who
normally receive high intravenous rates of isotonic saline.
Not only urine sodium concentration is almost always
significantly below 140 meq/L, but also the same is true
for stool, sweat, saliva and other bodily fluids ( table1 ) [10,
11] . More importantly, the urine sodium plus potassium
concentration is also usually below 140 meq/L.
On a quantitative basis, if the serum sodium concen-
tration increases from 140 to 141 meq/L, the sense of
thirst will make us drink water and subsequently lower
the serum sodium concentration to 140 meq/L again
( fig.8 ). Thus, it is easy to see that the only way to become
hypernatremic is to have lost the sense of thirst or access
to water. On the other hand, if serum sodium concentra-
tion goes from 140 to 139 meq/L, then vasopressin secre-
tion would be shut off and water excretion in the kidneys
will bring the serum sodium concentration back to 140
meq/L. Again, one can see that abnormalities in vaso-
pressin secretion are the main reason for hyponatremia,
whether it is appropriate or inappropriate.
Hypernatremia
The best way to understand the management of hyper-
natremia is to categorize it into 2 different circumstances
most commonly encountered in clinical medicine. The
147.9 meq of Na
move from
intracellular
to extracellular
Na = 14
K = 140
Cl = 4
osm = 300
H
2O = 22 l
Na = 142
K = 4.2
Cl = 108
osm = 300
H
2O = 18 l
ECF ICF
100 meq of KCl
98.6 meq intracellular
1.4 meq extracellular
Na = 7.5
K = 149.2
Cl = 1.8
osm = 305
H
2O = 21.3 l
Na = 144.6
K = 4.1
Cl = 106.6
osm = 305
H
2O = 18.7 l
ECF ICF
Fig. 7. Electrolyte concentrations before and after administration
of 100 meq of KCl. Assuming: (1) two-compartment model, (2)
ionic osmolality of 1.0, (3) complete osmotic equilibrium between
the compartments, (4) KCl distributes in intracellular and extra-
cellular fluids, (5) intercompartmental movement of Na with K
administration. Ignoring: (1) the Gibbs-Donnan effect, (2) non-
osmotically active Na and K.
Table 1. Sodium concentration of some bodily fluids
Bodily fluid Na concentration, meq/L
Urine 4060
Stool <50
Sweat 3656
Spit 15
Sam /Feizi Am J Nephrol 2012;36:97104 102
management is similar but slightly different in each of
these situations, emphasizing that in all of these cases the
patient has either lost the sense of thirst or is unable to
access water (i.e. intubated in the intensive care unit).
The typical patient in the first situation is the elderly
nursing home resident with dementia who has lost the
sense of thirst. The patient is brought to the emergency
room and has a serum sodium concentration of for in-
stance 170 meq/L, and there may or may not be coexisting
acute illnesses. The development of hypernatremia in this
patient has occurred slowly over many days and the treat-
ment should thus be gradual. In fact, if one assumes that
the patient drinks 450 ml of water a day and takes in 0.5
g/day of sodium and potassium each (patient is hardly
eating and drinking), makes 500 ml of urine/day (patient
is oliguric because of dehydration) which has a sodium
concentration of 40 meq/L and potassium concentration
of 30 meq/L and has total body water of 30 liter, one can
calculate that the development of hypernatremia has oc-
curred over 138 days (or almost 5 months). It is clear that
we should not correct this patients serum sodium con-
centration rapidly. In fact even the recommended correc-
tion of 10 meq/L/day may be too fast pending further
studies [12] . The latter patient is dehydrated and volume
depleted, and since sodium correction should be slow, it
is probably indicated to begin treatment with isotonic sa-
line and then switch to a less hypertonic solution depend-
ing on the response of serum sodium concentration. The
isotonic saline infusion can result in an increase in serum
sodium concentration depending on the urine sodium
and potassium concentration and the urinary volume
(another words the electrolyte free water excretion). In
our opinion, this patient should have his/her serum so-
dium concentration corrected very slowly and may not
necessarily need to go home with a normal serum sodium
concentration. On the other hand if a similar patient
drinks 200 ml of water a day and takes in 1 g/day of so-
dium and potassium each, makes 500 ml of urine/day
which has a sodium concentration of 40 meq/L and po-
tassium concentration of 30 meq/L and has total body
water of 30 liter, the calculation for development of hy-
pernatremia is over 12 days only. However, the former
situation is probably more frequent, as the latter situation
will require an abrupt change in mental status as the flu-
id intake is so low.
Next is the patient in the intensive care unit who has
developed acute renal failure [13] . He is intubated and on
mechanical ventilation and does not have free access to
water. At the beginning of his illness when he first devel-
oped sepsis and hypotension, he received many liters of
isotonic saline. At this time he had developed acute renal
failure and was not making much urine. His serum so-
dium concentration has remained normal despite having
an intake consisting mainly of isotonic saline because of
his low urine output (serum sodium concentration stays
at the normal level because he is not losing much hypo-
tonic fluids even though he is receiving isotonic fluids).
The hypernatremia arises when he starts recovering renal
function, making 34 liter a day of urine with low sodium
and potassium concentration. At this time, he is still re-
No
water
over
few
hours
A 1%
increase in
Na and
osmolality
Feel
thirsty
Drink
water
Na back
to
normal
Large
water
intake
Fall in serum
sodium and
osmolality
Shut off
ADH
secretion
Water
diuresis
Na back
to
normal
Fig. 8. Normal sodium balance.
Sodium Disorders Am J Nephrol 2012;36:97104 103
ceiving isotonic fluid for the most part, while losing many
liters of relatively electrolyte-free urine, leading to an in-
crease in serum sodium concentration. The serum sodi-
um concentration rises by 46 meq/L/day. For example,
this patient may be making 2 liter of urine with a urine
Na concentration of 32 meq/L and urine K concentration
of 45 meq/L. Assuming a starting serum sodium concen-
tration of 140 meq/L, the electrolyte-free water excretion
would be 0.9 liter. If the patient does not have any intake
and has no other output, then the serum sodium concen-
tration the next day would be 142.6 meq/L (the total body
water is 50 liter as the patient has anasarca from earlier
saline administration), [(140 meq/L ! 50 l) (45 meq/L
+ 32 meq/L) ! 2]/48 l.
This situation is probably by far more common than
the nursing home patient. The treatment of this patient
should consist of mainly free water and correction of se-
rum sodium quickly is not to be feared as hypernatremia
has developed over a few days too. This patient should
have as little sodium intake as possible as he has massive
amount of extra sodium already (even half isotonic saline
should be avoided). Also this patient is not volume de-
pleted as earlier he has received many liter of fluids with-
out making much urine and has extensive edema and of-
ten anasarca. The usual mistake in this case is not giving
enough free water. Four factors combine to cause the hy-
pernatremia in this patient. First, the patient is unable to
access water as he is intubated. In fact, a similar patient
who is not in the intensive care unit will not develop hy-
pernatremia as they would be able to drink water. Second,
this patient has renal failure (thus limiting the ability of
the kidneys to make concentrated urine). With normal
renal function, we are able to maximally concentrate our
urine and have large excretion of sodium and potassium
(people with normal renal function can increase urine
sodium concentration dramatically with saline adminis-
tration). Third, this patient is losing large amounts of flu-
id (at least 12 l/day). Most commonly, this occurs in the
setting of high urine output after recovery of tubular ne-
crosis, but can also happen with even large amounts of
diarrhea. Fourth, commonly this patient has a high se-
rum urea concentration which competes with sodium
and potassium for urinary excretion.
Conclusion
In summary, four factors combine to lead to hyperna-
tremia in a great majority of patients who develop this
disorder. First and foremost, the patient has lost the sense
of thirst or access to water. If this was not the case, the
patient would drink their way out of hypernatremia in
spite of the presence of other factors that may lead to hy-
pernatremia. Because of this factor, hypernatremia in the
hospital mostly occurs in patients in the ICU, intubated
on mechanical ventilation. Less commonly, confused pa-
tients on the regular hospital wards can get hyperna-
tremic. Second, there usually is an impaired ability to
concentrate the urine. Even though this happens in cases
of diabetes insipidus, it is much more common in the set-
ting of renal failure whether it is acute or chronic. If a
person with normal renal function and normal renal con-
centrating ability receives high rate of isotonic saline
(2 l/h) for 1 h, their urine osmolality would be in the order
of 600700 mosm/l, their urine sodium concentration in
the order of 110200 meq/L and their urine potassium
concentration would be in the order of 40 meq/L [14] .
Thus, they would not develop significant hypernatremia
despite making large amounts of urine even if there is an
impairment in the sense of thirst. Patients with kidney
disease are unable to concentrate their urine to 600700
mosm/l and in fact most often their urine osmolality is
only in the order of 300450 mosm/l [13] . This limits the
patients ability to excrete the extra sodium compared to
people with normal kidney function. Volume-depleted
patients, as happens in the nursing home patients, may
also be unable to excrete large sodium loads because of
the kidneys role in trying to correct the low volume state
by reabsorbing sodium (although this occurs much less
commonly). Third, the high serum urea concentration
leads to urea competing with sodium to be excreted in the
urine. We rather think of it as urea competing with elec-
trolytes than urea inducing osmotic diuresis, because
here the urine osmolality is limited by the renal failure.
Thus, even though the high urea excretion may lead to
more water excretion, it would not lead to more electro-
lyte excretion as the osmolality of the urine is fixed at
300450 mosm/l. One can think of it as: the high serum
sodium and urea concentrations lead them to compete to
get out into the urine which has limited ability to excrete
osmoles. Lastly, the large output, whether it is urine (most
commonly) or stool, leads to a large water excretion which
is not balanced by large sodium and potassium excretion.
Even though the larger the output the more likely it is that
hypernatremia will develop, increases in serum sodium
concentration can happen with outputs as low as one liter
per day.
Often, patients with hypernatremia in the intensive
care unit are not volume depleted despite having large
outputs as they have received large volumes of fluid ear-
Sam /Feizi Am J Nephrol 2012;36:97104 104
lier in the course of their illness and have had much less
output for many days [13] . Thus, hypernatremia often de-
velops in the setting of large outputs and not in the setting
of volume depletion.
Disclosure Statement
The authors have no conflicts of interest to declare.

References
1 Edelman IS, Leibman J, OMeara MP,
Birkenfeld LW: Interrelations between sodi-
um concentration, serum osmolality and to-
tal exchangeable sodium, total exchangeable
potassium and total body water. J Clin Invest
1958; 37: 12361256.
2 Adrogue HJ, Madias NE: Hypernatremia. N
Engl J Med 2000; 342: 14931499.
3 Anderson RJ, Chung HM, Kluge R, Schrier
RW: Hyponatremia: a prospective analysis of
its epidemiology and the pathogenetic role of
vasopressin. Ann Int Med 1985; 102: 164168.
4 Kurtz I, Nguyen MK: Evolving concepts in
the quantitative analysis of the determinants
of the plasma water sodium concentration
and the pathophysiology and treatment of
the dysnatremias. Kidney Int 2005; 68: 1982
1993.
5 Titze J, Shakibaei M, Schafflhuber M, Schul-
ze-Tanzil G, Porst M, Schwind KH, Dietsch
P, Hilgers KF: Glycosaminoglycan polymer-
ization may enable osmotically inactive Na
+

storage in the skin. Am J Physiol Circ Physi-
ol 2004: 287:H203H208.
6 Rose BD: New approach to disturbances in
the plasma sodium concentration. Am J Med
1986; 81: 10331040.
7 Nguyen MK, Kurtz I: New insights into the
pathophysiology of the dysnatremias: a qual-
itative analysis. Am J Physiol Renal Physiol
2004; 287:F172F180.
8 Nguyen MK, Kurtz I: Correction of hyper-
volaemic hypernatraemia by inducing nega-
tive Na
+
and K
+
balance in excess of negative
water balance: a new quantitative approach.
Nephrol Dial Transplant 2008; 23: 22232227.
9 Bhave G, Neilson EG: Body fluid dynamics:
back to the future. J Am Soc Nephrol 2011; 22:
21662181.
10 Takamata A, Yoshida T, Nishida N, Mori-
moto T: Relationship of osmotic inhibition
in thermoregulatory responses and sweat
sodium concentration in humans. Am J Phy-
siol Regul Integr Comp Physiol 2001;
280:R623R629.
11 Murphy MS, Sibal A, Mann JR: Persistent di-
arrhea and occult VIPomas in children. BMJ
2000; 320: 15241526.
12 Burrows FA, Shutack JG, Crone RK: Inap-
propriate secretion of antidiuretic hormone
in a postsurgical pediatric population. Crit
Care Med 1983; 11: 527531.
13 Sam R, Hart P, Haghighat R, Ing TS: Hyper-
volemic hypernatremia in patients recover-
ing from acute renal failure in the intensive
care unit. Clin Exp Nephrol 2012; 16: 136
146.
14 Lobo DN, Stanga Z, Alastair J, Simpson D,
Anderson JA, Rowlands BJ, Allison SP: Dilu-
tion and redistribution effects of rapid 2-litre
infusions of 0.9% (w/v) saline and 5% (w/v)
dextrose on haematological parameters and
serum biochemistry in normal subjects: a
double-blind crossover study. Clin Sci 2001;
101: 173179.