Clin. Cardiol.
14, 165-168 (1991)
Cardiac Conduction Defects Associated with Hyponatremia
M. MOUALLEM.
M.D., E. FRIEDMAN,
M.D., Y . SHEMESH, M.D..*
Departments of Internal Medicine E, and *Cardiology, Chaim Sheba Medical Center, Tel Hashomer and Sackler School of
Medicine, Tel Aviv University, Tel Aviv, Israel
Summary: Cardiac conduction defects have not been previ-
ously described in association with hyponatremia, although
in patients with congestive heart failure the frequency of
ventricular premature beats was found to correlate to the
severity of' hyponatremia. We describe three patients with
second-degree or complete atrioventricular (AV) block
which occurred during or shortly after an episode of se-
vere hyponatremia. The first had thiazide-induced
hyponatremia while on amiodamne. In the second, definite
etiology for hyponatremia which was associated with long-
standing polydipsia could not be established. The third had
ischemic heart disease and intermittent conversion of his
first-degree to second-degree AV block while hyponatremic
after diuretics use. Although it is usually difficult to single
out hyponatremia as the cause of conduction defects which
usually occur in the presence of cardiac disease, potent
medications or other electrolyte abnormalities, we suggest
that hyponatremia may play a role in the pathogenesis of
conduction defects in the diseased heart.
Key wards: heart block, hyponatremia
Introduction
Hyponatremia, the most common electrolyte disorder
in hospitalized patients is usually asymptomatic. I When
clinical manifestations of hyponatremia do occur they are
usually related to central nervous system dysfunction: con-
fusion, convulsions, coma, and death. The pathophysio-
Address for reprints:
Zvi Farlel, M.D.
Department of Internal Medicine E
Sheba Medical Center
Tel Hashomer 5262 I
Israel
Received: April 2, 1990
Accepted with revision: July I , 1990
M.D.. R. PAUZNER, M.D.,
H. MAYAN, z. FARFEL,M.D.
logical basis for these is serum hypo-osmolality which in
turn causes neuronal cell swelling and dysfunction.*
Clinical cardiac toxicity associated with hyponatremia
has not been described, although in patients with conges-
tive heart failure it was shown that the number of ventric-
ular premature beats (VPBs) correlated to the same degree
with the severity of either hyponatremia or hypokalemia.
In this report we describe three patients who developed
reversible cardiac conduction defects temporally associated
with hyponatremia or its correction.
Patient 1
A 75-year-old woman with hypertension and paroxys-
mal atrial fibrillation was treated by digoxin 0.25 mg and
amiodarone 200 mg daily. One week prior to admission
Kaluril (hydrochlorothiazide 50 mg, amiloride 5 mg) was
started. She was admitted because of progressive weak-
ness. On examination she was oriented. The pulse was
regular at 68/min, and blood pressure was 130/90 mmHg.
Cardiac examination was unremarkable and the rest of the
physical examination was normal. Electrocardiogram
showed first-degree atrioventricular (AV) block with P-
R interval of 0.24 s and complete left bundle-branch block
(CLBBB). A few hours later she became confused. Blood
pressure was unchanged and on ECG complete AV block
appeared, requiring insertion of a temporary pacemaker.
Pacemaker insertion did not alter mental status. Serum so-
dium concentration at that time was 120 mmol/l and potas-
sium was 5 mmol/l (Table I). When serum sodium reached
126 mmol/l, the patient became oriented and the ECG
showed 2: 1 AV block. A day later ECG showed normal
sinus rhythm and CLBBB. The CLBBB disappeared when
serum sodium level rose above 130 mmol/l. There was
no electrocardiographic or enzymatic evidence of my-
ocardial infarction. Serum digoxin concentration on ad-
mission was 0.3 ng/ml.
Comment
The patient presents a case of thiazide-induced
hyponatremia, which typically occurs in elderly women
166 Clin. Cardiol. Vol. 14. Fcbruary 1991

TABLEI Coursc of electrolyte and electrocardiographic changes in two patients

Hospitalization Serum sodium Serum potassium Serum

day (mmol/l) (mmol/l) bicarbonate (minol/l) ECC

Patient I
1 CLBBB, and first-degree
A V block (P-R 0.24S )
I 120 5 .O Completc AV block
2 117 5.3 18 Complete A V block
alternating with CLBBB
3 126 4.8 24 2:l A V block. CLBBB
4 131 5.1 22 Normal sinus rhythm
5 128 5.I Normal sinus rhythm
6 137 4.9 24 Normal sinus rhythm
7 14 I 5.3 Normal sinus rhythm
Patient 2
I 108 4.1 First-degree A V block
(P-R 0.44 S )
I29 4.7 20 P-R 0.34s
134 4.2 20 Wcnkebach type AV block
I39 4.2 16 Normal sinus rhythm
147 4.1 14 Nomial sinus rhythm
142 4.0 Normal sinus rhythm
140 4.2 20 Normal sinus rhythm
Ahbreviutions: CLBBB=complete left bundle-branch block, AV =atrioventricular.

a few days after thiazide therapy is ~ t a r t e d . The

~ unusual
appearance of conduction defects could not be explained
solely by a diseased conduction system, myocardial ische-
mia, or by amiodarone or digoxin therapy. The close tem-
poral association between the totally reversible conduc-
tion defect and hyponatremia strongly suggests that
hyponatremia played a role in the pathogenesis of the con-
duction defect.
Patient 2
A 33-year-old man was admitted because of confusion.
The patient sustained traumatic quadriplegia at age 20,
and because of an indwelling permanent urinary catheter
he used to drink large quantities of water. Five days prior
to admission he developed diffuse abdominal pains,
nausea, vomiting, and diarrhea, and later became drowsy
and dysarthric. On admission he was confused, his rectal
temperature was 33"C, blood pressure was 110/60 mmHg,
pulse rate was 52 beatdmin, hemoglobin concentration
was 90 g/l and serum sodium concentration was 108
mmol/l. ECG showed sinus bradycardia with first-degree
A V block, P-R interval of0.44 s, and J waves. Hyponatre-
mia was corrected by hypertonic saline, and serum sodi-
um concentration of 129 mmol/l was measured after 22
h (Table I). The patient became more oriented, his tem-
perature rose to 36"C, and the P-R intcrval shortened (0.34
s), however, 8 h later he became confused again. Serum
sodium, blood pressure, and body temperature wcrc all
normal. Wenkebach type second-dcgrce AV block ap-
peared on ECG. This electrocardiographic abnormality as
well as the sinus bradycardia and the first-degree AV block
spontaneously disappeared after 24 h (Table I). His con-
fusion lasted a few more days until complete recovcry ot
mental function. Serum cortisol and thyroxine concentra-
tions as well a5 chest x-rays were normal. The EECi
showed diffuse abnormality compatible with the severe
electrolyte disturbance. Cerebral computerized tomogra-
phy and examination of the ccrebrospinal fluid were nor-
mal. Thus, no definite cause for the hyponatrcniia could
be established. It was attributed partially at least to cx-
cessive drinking.'j
Comments
The cause of the hyponatremia in this paticnt was not
definitively established. Theoretically, a viral disease
could account for most or all clinical manifestations: my-
ocarditis, encephalitis, and concomitantly and indirectly-
hyponatremia.' No known virus could be dernonstratcd
in blood or cerebro spinal fluid (CSF) and thc EEG find-
ings argue against herpes encephalitis. Alternativcly, thc
hypothermia observed could have contributed to the ECG
changes. Hypothermia-associated ECG abnormalitics in-
clude bradycardia, atrial fibrillation, prolonged Q-T in-
terval, first-degree AV block, and the pathognomonic J
waves. 8 . 9 However, these changes are usually associatcd
 M. Mouallem et d . : Cardiac conduction defects and hyponatremia
with prolonged and profound hypothermia ( <30°C),and
to the best of our knowledge second-degree AV block has
never been associated with the mild degree of hypother-
mia observed in our patient. It is more plausible, there-
fore, that hyponatremia was involved in the conduction
system abnormalities, since the latter were noted on ad-
mission and improved upon correction of hyponatremia.
The delayed appearance of Wenkebach type AV block is
interesting, since it may represent a delayed type of dys-
function, analogous to the cerebral dysfunction encoun-
tered after correction of hyponatremia. l o
Patient 3
A 60-year-old man with ankylosing spondylitis, diabetes
mellitus, and ischemic cardiomyopathy (LVEF= 15%)
was known to have first-degree AV block, with P-R in-
terval of 0.28 s. His therapeutic regimen consisted of
furosemide, spironolactone, and digoxin 0.25 mg daily.
Five weeks before the last admission, he was hospital-
ized because of right heart failure with no evidence of
acute myocardial infarction. At that time ECG showed
P-R interval of 0.36 s and signs of old anterior myocardi-
al infarction. Serum digoxin concentration was 0.9 ng/ml,
serum sodium was 132 mmol/l, potassium 4.7 mmol/l,
and glucose 16.5 mmol/l. Digoxin was discontinued, and
chlorothiazide 0.5 g daily was added to the therapeutic
regime. One month later he was hospitalized because of
progressive weakness. There were no physical signs of
heart failure except for cachexia. Serum sodium was 116
mmolil, potassium 4.7 mmol/l, and glucose 18 mmol/l.
ECG showed P-R interval of 0.28 s and signs of old an-
terior myocardial infarction. Next day, at serum sodium
concentration of 1 18 mmol/l and potassium 5 mmol/l,
ECG showcd Mobitz type 2 second-degree AV block. Di-
uretics were discontinued and fluid intake was restricted
to 1,000 ml/day. During the next 14 days serum sodium
concentration ranged from 115 to 128 mmol/l. During
these days ECG showed Mobitz type 2 alternating with
Wenkcbitch type second-degree AV block. However, there
were days in which only first-degree AV block was record-
ed. During the rest of his hospital stay, when serum sodi-
um concentrations were between 125-133 mmol/l, no
second-degree AV block was recorded. Reappearance of
severe heart failure demanded reinstitution of diuretics.
Multiorgan failure developed and the patient died 34 days
after admission. Permission to autopsy was not granted.
Comment
This patient had severe ischemic cardiomyopathy and
ankylosing spondylitis, each of which can potentially cause
conduction abnormalities. Evidently, this patient had a dis-
eased conduction system, however, it is highly probable
that appearance of hyponatremia worsened the function
of the discascd conduction system and contributed to the
I67
development of second-degree AV block. This defect was
not documented in previous hospitalizations when he was
not hyponatremic, or in the last weeks of his life, when
he was only mildly hyponatremic. It is interesting that dur-
ing the period of moderate to severe hyponatremia, the
second-degree AV block was only intermittent. This may
further indicate that hyponatremia was not the only ab-
normality responsible for the conduction defect.
Discussion
Two of the patients described (Nos. 1 and 3) had com-
plete or second-degree AV block during hyponatremia.
Both patients had either a diseased heart (Patient 3) or were
receiving antiarrhythmic agents (Patient 1). In one patient
who did not have a known heart disease (Patient 2),
hyponatremia was associated with a first-degree AV block,
and only after correction of hyponatremia second-degree
AV block was observed. In all three cases hyponatremia
was severe, with serum sodium concentration of 108-
I17 mmol/l. Serum potassium concentrations and acid
base indices were both normal in all patients.
Hyponatremia has not previously been associated with
cardiac conduction defects, although co-occurrence of
hyponatremia and sinoatrial block was described in one
patient. I I In that patient both abnormalities may have been
independently induced by carbamazepine therapy which
is known to cause both hyponatremia'* and cardiac con-
duction defects. l 3
Theoretically, reduction of the extracellular concentra-
tion of sodium should slow cardiac pacemaker activity.
In animal experiments, only severe reduction of the sodi-
um concentration in the fluid perfusing the isolated heart
caused reduction in its contraction rate as well as in its
excitability and conduction velocity. l4 When dogs were
induced to develop wide QRS complexes, either through
hyperkalemia or quinidine administration (a situation ap-
parently simulating a diseased myocardial conduction sys-
tem), infusion of sodium salt solution resulted in a decrease
in the duration of the QRS complex.15 In the turtle, iso-
lated heart perfusion with low sodium solution made the
heart more sensitive to the myocardial effects of hyper-
kalemia as shown on electrocardiography. I 6 These obser-
vations raise the possibility that low sodium concentra-
tions can affect the diseased human heart, a condition that
cannot be accurately simulated by an animal model.
Documentation of the effects of various sodium con-
centrations on the human myocardial conduction system
are scarce. Garcia-Palmieri described four patients with
hyperkalemia and hyponatremia who had wide QRS com-
plexes. Three of the patients also had a moderate to se-
vere metabolic acidosis. The patients received hyperton-
ic saline with reversal of the ECG changes. This effect
can be either attributed to a change in serum potassium,
o r to a modification of the cardiac effects of the hyper-
kalemia. This observation may provide further evidence
I68 Clin. Cardiol. Vol. 14, February 1991
that hyponatremia may aggravate or cause cardiac con-
duction defects in a diseased heart.
In the specific hyponatremic patient, ascribing conduc-
tion defect to hyponatremia is usually complicated by other
coexisting conditions which may have a direct deleteri-
ous effect on the conduction system: primary cardiac dis-
ease, diuretic and antiarrhythmic drugs, or other electro-
lyte abnormalities (mainly hypokalemia and hyperkale-
mia). The clinical course of the three patients herein
described suggests that hyponatremia may play a role in
the pathogenesis of the observed atrioventricular conduc-
tion defects.
The association between cardiac conduction defects and
hyponatremia is far from established. Future clinical ex-
perience with increased index of suspicion will help to
establish the relation between these abnormalities further,
and to understand the mechanism of this phenomenon
better.
References
I. Anderson RJ, Chung H, Kluge R, Schrier RW: Hyponatreniia:
A prospective analysis of its epidemiology and the pathogenetic
role of vasopressin. Ann Intern Mrd 102, 164 (1985)
2. Norenberg MD, Leslie KO, Robertson AS: Association between
rise in seturn sodium and central pontine niyelinolysis. Ann Neu-
rol II, 128 (1982)
3. Dargie HJ, Cleland JGF, Leckie BJ, lnglih CG, East BW, Ford
I: Relation of arrhythmias and electrolyte abnormalities to sur-
vival in patients with severe chronic heart failure. Circulution
75 (suppl IV), 98 (1987)
4. Abramow M, Cogan E: Clinical aspects and pathophysiology
of diuretic induced hyponatremia. Ad\' Nc,plirol 13. I (1984)
5 . Friedman E, Shadel M , Halkin H, Farfel Z:Thiazidc-induced
hyponatremia: Reproducibility by single dose rcchallenge in-
cluding an analysis of pathogenesis. Ann Intcwi Mrd 110. 24
( 1989)
6. Leehey DJ, Picache AA, Robertson GL: Hyponatremia in quad-
riplegic patients. CIin Sci 75, 441 (1988)
7. Mouallem M, Friedman E, Rubinstein E: Inappropriate antidi-
uretic hormone secretion with infectious mononucleosis (let-
ter). N Engl J Med 3 II, 262 ( 1984)
8. Clements SD, Hurst JW: Diagnostic value of electrocardio-
graphic abnormalities in subjects accidentally exposed t o cold.
Am J Cardiol 29, 729 (1972)
9. Trevino A, Razi B, Beller BM: The characteristic electrocardi-
ogram of accidental hypothermia. Arch Intern Mod 127. 470
(1971)
10. Arieff A I: Hyponatremia, convulsions, respiratory arrest, and
permanent brain damage after elective surgery in healthy woni-
en. N Engl J Med 314, 1529 (1986)
II. Johannessen AC, Nielsen OA: Sino-atrial block, hyponatrae-
niia and urticaria caused by carbamazepinc. 0go.c.kr LUryyr 149,
376 (1987)
12. Ashton MG. Ball SG, Thomas TH, Lee MR: Water intoxicn-
tion associated with carbamazepine treatment. Br M e d J I, I134
( 1977)
13. Boesen F, Andersen EB. Jensen EK, Ladefogcd SD: Cardiac
conduction disturbances during carbamazepine therapy. k t o
Nrurol Sccmd 68, 49 (1983)
14. Trautwein W: Generation and conduction of impulses in the
heart as affected by drugs. Phartnucol Rrv 15, 277 (1963)
15. Surawicz B: Relationship between electrocardiogram and elec-
trolytes. An1 Heirrt J 73, 814 (1967)
16. Butcher WA, Wakim KC, Essex HE, Pruitt RD. Burchell HR:
The effect of changes in concentration of cations o n the clcc-
trocardiogram of the isolated perfused heart. Atn Hetrrt J 4 I,
801 (1952)
17. Garcia-Palmieri MR: Reversal of hyperkalemic cardiotoxicity
with hypertonic saline. Am Hrurt J 64, 483 (1962)