CLINICAL RESEARCH STUDY

Etiologic and Therapeutic Analysis in Patients

with Hypokalemic Nonperiodic Paralysis
Chih-Chien Sung, MD,a,b Chih-Jen Cheng, MD, PhD,a,b Wen-Fang Chiang, MD,a,c Tom Chau, MD,d Yu-Juei Hsu, MD, PhD,a
Sung-Sen Yang, MD, PhD,a,b Shih-Hua Lin, MDa,b
a
Division of Nephrology, Department of Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan;
b
Graduate Institute of Medical Science, National Defense Medical Center, Taipei, Taiwan; cDepartment of Medicine, Armed Forces Taoyuan
General Hospital, Taoyuan, Taiwan; dDepartment of Medicine, Providence St. Vincent Medical Center, Portland, Ore.

ABSTRACT

BACKGROUND: Hypokalemic nonperiodic paralysis represents a group of heterogeneous disorders with a

large potassium (K) decit. Rapid diagnosis of curable causes with appropriate treatment is challenging to
avoid the sequelae of hypokalemia. We prospectively analyzed the etiologies and therapeutic characteristics
of hypokalemic nonperiodic paralysis.
METHODS: Over an 8-year period, patients with hypokalemic nonperiodic paralysis were enrolled by
excluding those with hypokalemic periodic paralysis due to acute shift of K into cells. Blood and spot
urine samples were collected for the measurements of electrolytes, pH, and biochemistries. Intravenous
potassium chloride (KCl) at a rate of 10-20 mmol/h was administered until muscle strength recovered.
RESULTS: We had identied 58 patients with hypokalemic nonperiodic paralysis from 208 consecutive
patients with hypokalemic paralysis, and their average K concentration was 1.8  0.2 mmol/L. Among
patients with low urinary K excretion (n 17), chronic alcoholism, remote diuretic use, and anorexia/
bulimia nervosa were the most common causes. Among patients with high urinary K excretion (n 41)
and metabolic acidosis, renal tubular acidosis and chronic toluene abuse were the main causes, while
primary aldosteronism, Gitelman syndrome, and diuretics were the leading diagnoses with metabolic
alkalosis. The average KCl dose needed to restore muscle strength was 3.8  0.8 mmol/kg. Initial lower
plasma K, volume depletion, and high urinary K excretion were associated with higher recovery KCl
dosage. During therapy, patients with paradoxical hypokalemia (n 32) who required more KCl sup-
plementation than patients without (4.1  0.7 vs 3.4  0.7 mmol/kg, P < 0.001) often exhibited signi-
cantly higher plasma renin activity and received a higher volume of normal saline before its appearance.
CONCLUSIONS: Understanding the common etiologies of hypokalemic nonperiodic paralysis may aid in
early diagnosis. Patients with initial lower plasma K, renal K wasting, and hypovolemia required higher
recovery K dosage. Paradoxical hypokalemia is prone to develop in hypovolemic patients even during K
supplementation with volume repletion.
2014 Elsevier Inc. All rights reserved.  The American Journal of Medicine (2014) -, ---

KEYWORDS: Etiology; Hypokalemia; Paralysis; Treatment

Hypokalemia is a laboratory nding but can cause

Funding: This study was supported by grants from the Research Fund
of the Tri-Service General Hospital (TSGH-C101-114) and the Teh-Tzer
Study Group for the Human Medical Research Foundation.
Conict of Interest: None.
Authorship: All authors had access to the data and a role in writing the
manuscript.
Requests for reprints should be addressed to: Shih-Hua Lin, MD, Di-
vision of Nephrology, Department of Medicine, Tri-Service General Hos-
pital, No 325, Section 2, Cheng-Kung Road, Neihu 114, Taipei, Taiwan.
E-mail address: l521116@gmail.com
morbidity and mortality due to critical complications such as
cardiac arrhythmias and respiratory failure.1,2 In some cir-
cumstances, severe hypokalemia can cause neuromuscular
paralysis (so-called hypokalemic paralysis), a potentially
reversible medical emergency with unique diagnostic and
therapeutic challenges.3,4 Based on the pathophysiology, it
can be divided into 2 groups: hypokalemic periodic paral-
ysis due to an acute shift of potassium (K) into cells, and

0002-9343/$ -see front matter 2014 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.amjmed.2014.09.027
2 The American Journal of Medicine, Vol -, No -, - 2014

hypokalemic nonperiodic paralysis resulting from a large Denitions

decit of K.5,6 Due to often indistinguishable clinical Hypokalemic periodic paralysis was diagnosed by acute
features, we had shown previously that an assessment muscle weakness with inability to ambulate accompanied by
of urinary K excretion and blood acid-base status coupled hypokalemia, low urinary K excretion, and the presence of
with a detailed history and physical examination help thyrotoxic periodic paralysis, familial periodic paralysis, or
separate the 2 groups.7 Among the etiologies of hypokale- sporadic periodic paralysis. The rest were hypokalemic
mic periodic paralysis, the most nonperiodic paralysis, which is a
common are thyrotoxic periodic more appropriate term than non-
paralysis8-13 and sporadic periodic CLINICAL SIGNIFICANCE
hypokalemic periodic paralysis
paralysis14 in Asia and familial  Approximately 25% of cases of hypoka- published previously. We distin-
periodic paralysis in Western lemic paralysis were secondary to hypo- guish hypokalemic periodic paral-
countries.15,16 Within hypokale- ysis from hypokalemic nonperiodic
kalemic nonperiodic paralysis with
mic nonperiodic paralysis, the paralysis based on history such as a
chronic alcoholism, recent/remote
etiologies are complex and chal- positive family history, clinical
diuretic use, anorexia/bulimia nervosa,
lenging to the clinician hoping to course, urinary K excretion, and
make an early diagnosis. The renal tubular acidosis, primary aldoste-
acid-base status.7
correct diagnosis of the underlying ronism, and Gitelman syndrome being
hypokalemia avoids recurrence the most common.
and missing any potentially cur-  Hypokalemic patients with initial lower METHODS
able and treatable disorders. All patients received a standard
plasma potassium (K ), volume deple-
The strategy and dosage of K 12-lead electrocardiogram. Blood
tion, and renal K wasting required much
supplementation should be guided pressure and electrocardiogram
by the early and correct diagnosis higher recovery K dosage.
monitoring were recorded. Muscle
of hypokalemic nonperiodic pa-  Hypokalemic patients with hypovolemia strength graded on a 5-point scale,
ralysis. For patients with hy- are often prone to develop paradoxical venous blood and spot urine
pokalemic periodic paralysis, samples were obtained at the rst
hypokalemia during K supplementation
potassium chloride (KCl) should with aggressive volume repletion. visit in the emergency department,
be infused intravenously to and then every 1-2 hours before
quickly reverse hypokalemia, but recovery. Blood gases were
should be minimized to avoid rebound hyperkalemia in the measured by ABL 510 (Radiometer, Copenhagen,
recovery phase. In hypokalemic nonperiodic paralysis Denmark). Plasma and urine biochemical studies were
patients, intravenous KCl infusion is usually followed by determined by automated methods (AU 5000 chemistry
chronic oral KCl tablets to replete the large decit of total analyzer; Olympus, Tokyo, Japan) and supine plasma renin
body K. However, the therapeutic course and factors that activity and aldosterone level were also measured.
inuence the effectiveness of K supplementation in Intravenous KCl supplementation was initiated at a rate
patients with hypokalemic nonperiodic paralysis have of 10-20 mmol/h, but infused rate could be increased to
not been well characterized. In this study, we analyzed 30 mmol/h through a central venous catheter if the patient
the etiologies and therapeutic courses of 58 consecutive developed ventricular arrhythmias, respiratory failures, or
patients with hypokalemic nonperiodic paralysis over an a fall in plasma K concentration during therapy. The rate
8-year span at a single medical center to improve our ability of KCl supplementation was lowered to <10 mmol/h
to diagnose and optimize treatment for these critical when muscle strength recovered enough for the patient to
patients. ambulate. Patients with coexisting hypomagnesemia
received intravenous elemental magnesium 400-600 mg.
MATERIALS AND METHODS The total amount of KCl administered up to the time of
recovery was recorded. The recovery time was dened as
Study Subjects the time period from the start of KCl therapy to the recovery
The study protocol was approved by the Ethics Com- of muscle strength. Nadir plasma K was the lowest plasma
mittee on Human Studies at Tri-Service General Hospital. K concentration during KCl therapy. Paradoxical hypo-
As part of this study, the history, urine and plasma kalemia was dened as a fall of plasma K concentration
samples, and hospital course of patients admitted with 0.1 mmol/L during KCl therapy.17

hypokalemia (plasma K level <3.5 mmol/L) and muscle
paralysis (Medical Research Council grade 3) have been
collected prospectively from January 2005 through Statistical Analysis
December 2012. Patients who had had previous hypoka- Results are expressed as mean  standard deviation. The
lemic attacks or currently were receiving KCl treatment nonparametric Mann-Whitney U rank-sum test was used for
or been diagnosed as hypokalemic periodic paralysis were continuous data to compare the differences in variables
excluded. between 2 subgroups were not distributed normally.
Sung et al Etiology and Management of Hypokalemic Nonperiodic Paralysis 3

Categorical variables were compared using the chi-squared

Table 1 Characteristics of Patients with Hypokalemic Non-
test. One-way analysis of variance was used to compare periodic Paralysis (n 58)
independent variables among groups. Statistical signicance
Age (years) 40.7  18.7
was dened as P value < .05.
Sex (male/female) 44/14
Body mass index (kg/m2) 23.1  4.1
RESULTS Onset of paralysis, n (%)
Morning 31 (53.4%)
Patients Characteristics in Hypokalemic Afternoon 17 (29.3%)
Nonperiodic Paralysis Evening 10 (17.2%)
Muscle strength (MRC grade)
As shown in Figure 1, we had identied 58 patients with
Upper extremities 3.2  0.8
hypokalemic nonperiodic paralysis and 150 patients with Lower extremities 1.3  0.8
hypokalemic periodic paralysis consisting of thyrotoxic Plasma K (mmol/L) 1.8  0.2
periodic paralysis (n 107) and sporadic periodic Electrocardiography ndings, n (%)
paralysis (n 43) from 208 patients with hypokalemic Prominent U waves 46 (81.0%)
paralysis. The average initial plasma K concentration in Flatten T waves 37 (63.8%)
these patients with hypokalemic nonperiodic paralysis was ST depression 32 (55.2%)
1.8  0.2 mmol/L. Other characteristics were shown in First-degree atrioventricular block 18 (31.0%)
Table 1. Hypophosphatemia was found in 27 patients Premature ventricular complexes 10 (17.2%)
(46.5%), followed by hypocalcemia (34.5%) and MRC Medical Research Council.
hypomagnesemia (31.0%).

Cl excretion. Urinary high Na but low Cl excretion was

Etiologic Analysis in Hypokalemic Nonperiodic
Paralysis
Patients with Low Urinary KD Excretion. Seventeen
patients (29.3%) had low urinary K excretion (urine
K/creatinine ratio 1.1  0.6 mmol/mmol) and metabolic
alkalosis or acidosis (Table 2).
Two patients with metabolic acidosis (bicarbonate 17.3
 1.0 mmol/L) also had urinary low sodium (Na) and high
chloride (Cl) excretion. One had a distant history of
toluene abuse for recreational purposes and the other had
used acetazolamide to treat seizure.
There were 15 patients with metabolic alkalosis (bicar-
bonate 32.7  6.2 mmol/L). Patients with low urinary Na
and Cl excretion were either chronic alcoholics (n 6)
or remote users of loop and thiazide diuretics (n 4),
despite one alcoholic patient with high urinary Na and
found in 2 patients with anorexia/bulimia nervosa. Urinary
low Na but high Cl excretion was found in one patient
with chronic laxative abuse and one profound diarrhea.
Most of these etiologies (73%) were related to the gastro-
intestinal disorders combined with hypocalcemia (52.9%),
hypophosphatemia (41.2%), and hypomagnesemia (23.5%).
Patients with High Urinary KD Excretion. Forty-one
patients (70.7%) had high urinary K excretion (urine K/
creatinine ratio 4.9  2.8 mmol/mmol) and metabolic
acidosis or alkalosis (Table 2).
The etiologies in the 11 patients with hyperchloremic
metabolic acidosis (bicarbonate 15.6  4.8 mmol/L)
included renal tubular acidosis (n 8), recent toluene abuse
with high urine ammonium excretion18 (n 2), and ureteral
diversion (n 1). The causes of renal tubular acidosis

Hypokalemic paralysis
(n=208)

Hypokalemic periodic paralysis Hypokalemic non-periodic paralysis

(n=150) (n=58)

Thyrotoxic Sporadic Low potassium High potassium

periodic periodic excretion excretion
paralysis paralysis (n=17) (n=41)
(n=107) (n=43)

Figure 1 The causes and number in patients with hypokalemic paralysis in this study.
4 The American Journal of Medicine, Vol -, No -, - 2014

loop diuretics for weight reduction occurred in 4 patients

Table 2 Etiologies in Patients with Hypokalemic Nonperiodic
Paralysis (n 58) and 1 patient, respectively.
Hypocalcemia (54.5%) and hypophosphatemia (48.8%)
Age Sex were the most common accompanying abnormality, espe-
Number (Years) (M:F) cially in patients with renal tubular acidosis. Hypomagne-
Low urinary K excretion rate 17 42.5  6.1 15:2 semia was a typical laboratory nding in patients with
Hyperchloremic metabolic 2 Gitelman syndrome or thiazide diuretics.
acidosis
Remote toluene abuse 1 45 1:0
Remote acetazolamide use 1 35 1:0 Analysis of Therapeutic Characteristics
Hypochloremic metabolic 15
KCl Supplementation. The average recovery time and
alkalosis
Chronic alcoholism 7 46.7  8.2 7:0
recovery plasma K concentration amongst all patients
Remote loop diuretics & 4 60.8  24.9 3:1 were 13.0  3.6 hours and 2.6  0.2 mmol/L, respectively.
thiazides use The amount of KCl administered to achieve recovery was
Anorexia/bulimia nervosa 2 19.5  0.7 2:0 243  75 mmol (range 120-500 mmol, 3.8  0.8 mmol/
Profound diarrhea 1 57 1:0 kg). Patients with high urinary K excretion required
Chronic laxative use 1 34 0:1 higher recovery KCl dosage than those with low urinary
High urinary K excretion rate 41 33.5  4.2 29:12 K excretion (4.0  0.8 vs 3.2  0.5 mmol/kg, P < .05).
Hyperchloremic metabolic 11 Patients in the top quartile of recovery KCl dosage (n 15,
acidosis 336.0  59.6 mmol) had signicantly lower plasma K
Renal tubular acidosis 8 35.0  19.1 2:6 (1.8  0.3 vs 2.0  0.2 mmol/L, P < .05), higher blood
Recent toluene abuse 2 35.5  0.7 2:0
urea nitrogen (19.4  6.8 vs 9.8  0.2 mg/dL, P < .05),
Ureteral diversion 1 47 1:0
Hypochloremic metabolic 30
creatinine (1.16  0.31 vs 0.81  0.19 mg/dL, P < .05),
alkalosis and plasma renin activity (23.8  26.3 vs 6.9  8.3 ng/mL/
High blood pressure 11 h, P < .05) than those in the bottom quartile (n 15, 161.3
Primary aldosteronism 9 44.5  12.4 6:3  19.2 mmol) (Supplementary Table, available online).
Chronic licorice ingestion 1 40 0:1 Higher blood urea nitrogen and creatinine were primarily
Ectopic 1 60 1:0 caused by volume depletion because both rapidly
adrenocorticotropin decreased after saline infusion. Other factors such as
hormone syndrome metabolic alkalosis or acidosis, plasma Na, total calcium,
Normal blood pressure 19 magnesium, albumin, and phosphate levels did not affect
Gitelman syndrome 11 21.1  10.1 11:0 the total recovery KCl dosage.
Classic Bartter syndrome 1 20 1:0
Recent loop diuretics & 5 67.8  17.6 4:1
thiazides use
Patients with Paradoxical Hypokalemia. We noted that
Bulimia nervosa 2 20.0  2.8 1:1 32 patients developed paradoxical hypokalemia, in which
10 had low urinary K excretion and 22 had high urinary
K excretion. The nadir plasma K concentration (from
1.8  0.3 mmol/L to 1.6  0.2 mmol/L) was detected at a
included primary Sjgrens syndrome (n 7) and Fanconi median time of 2 hours (2.4  1.0 hours) into KCl ther-
syndrome (n 1). The patient with ureteral diversion had a apy. These patients required signicantly higher recovery
ureterosigmoidostomy created for neurogenic bladder.19 KCl dosage (4.1  0.7 vs 3.4  0.7 mmol/kg, P < .001)
Thirty patients with hypochloremic metabolic alkalosis than those without paradoxical hypokalemia in both low
(bicarbonate 33.7  5.6 mmol/L) could be further divided (3.4  0.5 vs 2.9  0.4 mmol/kg, P < .05) and high (4.4
into those with high (n 11) and normal (n 19) blood  0.6 vs 3.5  0.7 mmol/kg, P < .05) urinary K
pressure. Among patients with high blood pressure (164  excretion groups (Figure 2). These patients had a male
26/84  15 mm Hg), primary aldosteronism (n 9), predominance, higher plasma renin activity (Table 3).
chronic licorice ingestion (n 1), and ectopic adrenocor- After excluding patients with primary aldosteronism,
ticotropin hormone syndrome (n 1) were the underlying whose plasma renin activity and aldosterone level are
causes. not correlated with the volume status, patients with
For patients with normal blood pressure, the common paradoxical hypokalemia had a higher plasma renin
etiologies included Gitelman (n 11) or classic Bartter activity (21.1  21.4 vs 4.2  2.4 ng/mL/h, P < .05)
syndrome (n 1), recent diuretic use (n 5), and bulimia and plasma aldosterone level (301.8  460.8 vs 121.4 
nervosa (n 2). Gitelman syndrome was documented by 68.4 ng/mL/h, P < .05). In addition, they received a
homozygous or compound heterozygous mutations in the signicantly higher volume of normal saline infused in
SLC12A3 gene.20,21 The patient with classic Bartter syn- the rst 3 hours (807  100 vs 630  29 mL, P < .05).
drome had a homozygous 786delG mutation in the Seven of them developed new premature ventricular
CLCNKB gene. Thiazides prescribed for hypertension and contractions during therapy.
Sung et al Etiology and Management of Hypokalemic Nonperiodic Paralysis 5

Figure 2 Recovery KCl dosage to restore muscle strength in (A) all hypokalemic nonperiodic paralysis patients
(n 58), (B) hypokalemic nonperiodic paralysis patients with low K excretion (n 17), and (C) hypokalemic
nonperiodic paralysis patients with high K excretion (n 41). The shaded column denotes patients with para-
doxical hypokalemia (PH) and the white column, those without. *Denotes P < .05 and **P < .001.

DISCUSSION patients with low urinary K excretion, while renal tubular

This is the rst prospective study to analyze the etiologies acidosis, primary aldosteronism, recent diuretic use, and
and characterize the therapeutic courses of patients with Gitelman syndrome were the most common in patients with
hypokalemic nonperiodic paralysis. Among a diverse set of high urinary K excretion. Initial presentations of lower
etiologies, chronic alcoholism, remote diuretic use, and plasma K, volume depletion, and renal K wasting were
anorexia/bulimia nervosa were the common causes in associated with higher recovery KCl dosage. Hypovolemia
patients are prone to develop paradoxical hypokalemia
during K supplementation in saline infusion.
Table 3 Clinical and Laboratory Findings on Admission in In cases of hypokalemic paralysis, assessment of urinary
Hypokalemic Nonperiodic Paralysis Patients Without and With K excretion can separate renal K wasting from nonrenal
Paradoxical Hypokalemia K disorders.7 In emergent situations, when 24-hour urine
Without With collection is impractical, the spot urine K/creatinine can be
(n 26) (n 32) a valuable index of urinary K excretion.5 We found that
Age (years) 38.1  16.9 42.7  20.0
approximately one-third of our patients had low urinary K
Sex (male:female) 16:10 27:5 excretion. Aside from disorders involving the gastrointes-
Body mass index (kg/m2) 22.6  3.8 23.6  4.3 tinal tract (chronic alcoholism, anorexia/bulimia nervosa,
Recovery KCl dosage (mmol/kg) 3.4  0.7 4.1  0.7* profound diarrhea, and chronic laxative), these also included
Recovery time (hours) 11.8  3.4 13.8  4.1* former renal K wasting disorders, such as remote di-
Systolic blood pressure (mm Hg) 134  30 127  16 uretics and toluene abuse. An assessment of urine Na, Cl,
Heart rate (beats per minute) 83  15 80  17 and blood acid-base status helped distinguish gastrointes-
Hemoglobin (g/dL) 13.4  2.5 14.4  2.5 tinal disorders from nongastrointestinal former Na and
Bicarbonate (22-26 mmol/L) 27.9  8.7 29.3  9.0 K losing disorders.22,23
Na (135-142 mmol/L) 137.5  4.3 137.9  5.7 Approximately two-thirds of our patients had high
Initial K (3.5-5.0 mmol/L) 1.9  0.3 1.8  0.2
urinary K excretion. One-fourth of them had concomitant
Nadir K (3.5-5.0 mmol/L) 1.9  0.3 1.6  0.2*
Recovery K 2.7  0.2 2.6  0.2
hyperchloremic metabolic acidosis, caused by renal
(3.5-5.0 mmol/L) tubular acidosis, recent glue snifng, and ureteral diver-
Cl (98-106 mmol/L) 100.6  11.1 98.3  10.5 sion. Consistent with previous reports, most of our
Phosphate (2.6-4.5 mg/dL) 2.76  0.76 2.75  0.82 patients with renal tubular acidosis were female and had
Total calcium 8.8  0.9 8.7  1.2 autoimmune Sjgrens syndrome.24 Three-fourths of high
(8.4-10.2 mg/dL) urinary K excretion with hypochloremic metabolic
Magnesium (1.7-2.55 mg/dL) 1.98  0.37 1.96  0.45 alkalosis comprised a diverse set of etiologies. The
Albumin (3.5-5.7 g/dL) 3.9  0.6 4.0  0.5 patients with high blood pressure had mineralocorticoid
Blood urea nitrogen 13.9  6.3 15.7  7.8 excess secondary to primary aldosteronism, chronic lico-
(7-20 mg/dL) rice ingestion, and ectopic adrenocorticotropin hormone
Creatinine (0.7-1.2 mg/dL) 1.03  0.33 1.11  0.78
syndrome.25,26 The others with normal to low blood
Plasma renin activity 3.7  2.6 18.6  21.1*
pressure had renal salt wasting, and therefore enhanced
(0.1-2.0 ng/mL/h)
Plasma aldosterone level 227.7  256.2 267.8  439.4 urinary K and hydrogen excretion, due to Gitelman
(12-150 pg/mL) and Bartter syndromes, recent diuretic use, and bulimia
nervosa. Among these diseases, Gitelman syndrome was
*Denotes P < .05.
the most common.21,27,28
6 The American Journal of Medicine, Vol -, No -, - 2014

Other divalent abnormalities were also commonly
observed in our patients and may help identify the under-
lying cause. Hypophosphatemia was the most common
electrolyte abnormality, followed by hypocalcemia and hy-
pomagnesemia. Hypophosphatemia may be a consequence
of hypokalemia-related renal phosphate wasting and thus
not specic to any individual disease.29-31 However, coex-
isting with hypomagnesemia or hypocalcemia accompanied
by low urine magnesium or calcium are important clues for
gastrointestinal disorders.32-35 Signicant hypocalcemia and
hypophosphatemia can sometimes be seen in patients with
hyperchloremic metabolic acidosis. Metabolic acidosis per
se inhibits the synthesis of vitamin D and affects parathyroid
hormone, thereby reducing absorption of calcium and
phosphate in the gastrointestinal tract and kidneys.36
Distinct nding of hypomagnesemia with renal magne-
sium wasting and hypocalciuria hints at thiazide use or
Gitelman syndrome.
In terms of K therapy, patients with hypokalemic
nonperiodic paralysis require higher doses of KCl (3.8
mmol/kg) to recover muscle strength than those with hy-
pokalemia periodic paralysis (1 mmol/kg) whose
extracellular K ions are not lost in urine or feces but shifted
into cells.17 Among patients with hypokalemic nonperiodic
paralysis, those with high urinary K excretion needed
larger recovery KCl dosage than others with low urinary K
excretion. This could be partially explained by continued
wasting of the administered KCl in the urine while the
former renal and extrarenal patients simply had a K
decit in need of replacement. In addition, patients in the top
quartile of recovery KCl dosage had initial lower plasma K
and volume depletion as reected by higher plasma renin
activity and a rapid decline in plasma urea nitrogen and
creatinine than those in the bottom quartile.
Notably, paradoxical hypokalemia was observed in
approximately 50% of our patients. Although this inter-
esting phenomenon has been found in a few reports of
thyrotoxic periodic paralysis,17,37,38 it has not been
emphasized and claried in patients with hypokalemic
nonperiodic paralysis. In a thought experiment using the
results of our study (Figure 3), KCl replacement at 10
mmol/h over 2 hours (mean time to nadir) should result in
a net gain of 18 mmol, assuming urine K excretion is 1
mmol/h and there are no other sources of K loss. This

1L NaCl 140 mmol

Infused
and KCl 20 mmol

20 mmol K+ 18 mmol K+
[K+] 120 mmol/L [K+] 2.0 mmol/L [K+] 121 mmol/L [K+] 1.7 mmol/L
3Na+ 3Na+
- 2 K+
+
2 K+

1.6 mmol K+

ICF 24 L ECF 11 L ICF 24 L ECF 12 L

RMP K+ renal RMP
2 mmol
-108 mV excretion -116 mV

Volume depletion Volume repletion

Figure 3 The possible mechanisms for paradoxical hypokalemia in hypovolemic and hypokalemic patients
treated with volume repletion and KCl. For a 60-kg patient with a paradoxical reduction of plasma K level from 2.0 to
1.7 mEq/L after intravenous infusion of 1 L saline (w140 mmol NaCl) plus 20 mmol KCl into extracellular uid (ECF)
in 2 hours, a net 1.6 mmol-K is shifted into intracellular uid (ICF)
Total ECF K change: 12 L  1:7 mmol l  11 L  2 l
mmol
1:6 mmol
Assuming the urine K excretion rate is 1 mmol/h, the external balance (input  output) of K in this case is

18 mmol. To reach the net K loss of 1.6 mmol in ECF, total 19.6 mmol K is shifted into intracellular uid (ICF).
In addition to the initial predisposing event, volume repletion may ameliorate the a-adrenergic inhibition
on Na-K-ATPase and in turn further enhance K uptake into ICF. The resting membrane potential (RMP) is estimated
by Nernst equation.
Sung et al Etiology and Management of Hypokalemic Nonperiodic Paralysis
amount of net K should raise plasma K concentration in
the extracellular space. Therefore, a falling plasma K
concentration during KCl therapy indicates signicant
forces driving extracellular and infused K into cells.
We found that these patients with paradoxical hypoka-
lemia had signicant volume depletion, as evidenced by a
much higher plasma renin activity, and received a higher
volume of normal saline. It has been shown that volume
depletion can increase a-adrenergic activity, which inhibits
the activity of Na-K-ATPase through the suppression of
insulin secretion and thus causes K to exit from cells.17
These patients needed to receive more NaCl infusion
before its appearance. KCl administration also causes K
entry into cells in exchange with exiting intracellular Na
and helps expand extracellular volume. We suggest that the
increased volume effect of KCl and NaCl may diminish a-
adrenergic activity, releasing the inhibition of Na-K-
ATPase activity, which leads to the inux of K into
cells and a fall in plasma K (Figure 3). This scenario
resembles the hypokalemia that can develop during
volume expansion in diabetic patients with diabetic
ketoacidosis and initial hyperkalemia.39,40 Total body
decit of K is even greater than that suggested by the
initial plasma K concentration in patients with concomi-
tant hypovolemia as shown in these patients with para-
doxical hypokalemia.
There were some limitations of this study. First, we did
not obtain history about dietary K intake although most
patients had very little oral intake during the attack. Second,
we did not measure total urinary K excretion and overall
K balance (K input  K output) during therapy. Third,
we did not evaluate the effect of hypomagnesemia or hy-
pocalcemia that had been treated once identied on the
dosage of K supplementation.41
In conclusion, understanding the common etiologies of
hypokalemic nonperiodic paralysis may aid in its early
diagnosis. Patients with initial lower plasma K, volume
depletion, and renal K wasting required much higher re-
covery K doses to reach recovery. Furthermore, hypo-
volemic patients are often prone to develop paradoxical
hypokalemia, which remains difcult to identify clinically
and needs to be monitored closely during therapy. Our
ndings can also apply to nonparalytic patients with chronic
hypokalemia due to a K decit.
ACKNOWLEDGMENT
This study was supported by grants from the Research Fund
of the Tri-Service General Hospital (TSGH-C101-114) and
from the Teh-Tzer Study Group for the Human Medical
Research Foundation.
References
1. Halperin ML, Kamel KS. Potassium. Lancet. 1998;352:135-140.
2. Kamel KS, Quaggin S, Scheich A, Halperin ML. Disorders of potas-
sium homeostasis: an approach based on pathophysiology. Am J
Kidney Dis. 1994;24:597-613.
7
3. Ahlawat SK, Sachdev A. Hypokalaemic paralysis. Postgrad Med J.
1999;75:193-197.
4. Stedwell RE, Allen KM, Binder LS. Hypokalemic paralyses: a review
of the etiologies, pathophysiology, presentation, and therapy. Am J
Emerg Med. 1992;10:143-148.
5. Lin SH, Chiu JS, Hsu CW, Chau T. A simple and rapid approach to
hypokalemic paralysis. Am J Emerg Med. 2003;21:487-491.
6. Lin SH, Lin YF, Halperin ML. Hypokalaemia and paralysis. QJM.
2001;94:133-139.
7. Lin SH, Lin YF, Chen DT, Chu P, Hsu CW, Halperin ML. Laboratory
tests to determine the cause of hypokalemia and paralysis. Arch Intern
Med. 2004;164:1561-1566.
8. Ko GT, Chow CC, Yeung VT, Chan HH, Li JK, Cockram CS. Thyro-
toxic periodic paralysis in a Chinese population. QJM. 1996;89:463-468.
9. McFadzean AJ, Yeung R. Periodic paralysis complicating thyrotoxi-
cosis in Chinese. Br Med J. 1967;1:451-455.
10. Okinaka S, Shizume K, Iino S, et al. The association of periodic
paralysis and hyperthyroidism in Japan. J Clin Endocrinol Metab.
1957;17:1454-1459.
11. Kelley DE, Gharib H, Kennedy FP, Duda RJ Jr, McManis PG.
Thyrotoxic periodic paralysis. Report of 10 cases and review of elec-
tromyographic ndings. Arch Intern Med. 1989;149:2597-2600.
12. Ober KP. Thyrotoxic periodic paralysis in the United States. Report of
7 cases and review of the literature. Medicine. 1992;71:109-120.
13. Chang CC, Cheng CJ, Sung CC, et al. A 10-year analysis of thyrotoxic
periodic paralysis in 135 patients: focus on symptomatology and
precipitants. Eur J Endocrinol. 2013;169:529-536.
14. Sung CC, Cheng CJ, Lo YF, et al. Genotype and phenotype analysis of
patients with sporadic periodic paralysis. Am J Med Sci. 2012;343:
281-285.
15. Venance SL, Cannon SC, Fialho D, et al. The primary periodic pa-
ralyses: diagnosis, pathogenesis and treatment. Brain. 2006;129:8-17.
16. Matthews E, Labrum R, Sweeney MG, et al. Voltage sensor charge
loss accounts for most cases of hypokalemic periodic paralysis.
Neurology. 2009;72:1544-1547.
17. Shiang JC, Cheng CJ, Tsai MK, et al. Therapeutic analysis in Chinese
patients with thyrotoxic periodic paralysis over 6 years. Eur J Endo-
crinol. 2009;161:911-916.
18. Carlisle EJ, Donnelly SM, Vasuvattakul S, Kamel KS, Tobe S,
Halperin ML. Glue-snifng and distal renal tubular acidosis: sticking to
the facts. J Am Soc Nephrol. 1991;1:1019-1027.
19. Cruz DN, Huot SJ. Metabolic complications of urinary diversions: an
overview. Am J Med. 1997;102:477-484.
20. Lo YF, Nozu K, Iijima K, et al. Recurrent deep intronic mutations in
the SLC12A3 gene responsible for Gitelmans syndrome. Clin J Am
Soc Nephrol. 2011;6:630-639.
21. Tseng MH, Yang SS, Hsu YJ, et al. Genotype, phenotype, and follow-
up in Taiwanese patients with salt-losing tubulopathy associated with
SLC12A3 mutation. J Clin Endocrinol Metab. 2012;97:E1478-E1482.
22. Kamel KS, Ethier J, Levin A, Halperin ML. Hypokalemia in the
beautiful people. Am J Med. 1990;88:534-536.
23. Groeneveld JH, Sijpkens YW, Lin SH, Davids MR, Halperin ML. An
approach to the patient with severe hypokalaemia: the potassium quiz.
QJM. 2005;98:305-316.
24. Simpson DP. Control of hydrogen ion homeostasis and renal acidosis.
Medicine. 1971;50:503-541.
25. Lin SH, Chau T. A puzzling cause of hypokalaemia. Lancet. 2002;360:
224.
26. Lin CS, Yao NS, Cheng MF, Lin SH. Ectopic ACTH syndrome
associated with large-cell neuroendocrine carcinoma of the lung. Am J
Med Sci. 2007;334:487-489.
27. Cheng NL, Kao MC, Hsu YD, Lin SH. Novel thiazide-sensitive Na-Cl
cotransporter mutation in a Chinese patient with Gitelmans syndrome
presenting as hypokalaemic paralysis. Nephrol Dial Transplant.
2003;18:1005-1008.
28. Lin SH, Shiang JC, Huang CC, Yang SS, Hsu YJ, Cheng CJ.
Phenotype and genotype analysis in Chinese patients with Gitelmans
syndrome. J Clin Endocrinol Metab. 2005;90:2500-2507.
8 The American Journal of Medicine, Vol -, No -,
29. Anderson DC, Peters TJ, Stewart WK. Association of hypokalaemia
and hypophosphataemia. Br Med J. 1969;4:402-403.
30. Felsenfeld AJ, Levine BS. Approach to treatment of hypo-
phosphatemia. Am J Kidney Dis. 2012;60:655-661.
31. Mahler RF, Stanbury SW. Potassium-losing renal disease; renal and
metabolic observations on a patient sustaining renal wastage of po-
tassium. Q J Med. 1956;25:21-52.
32. De Marchi S, Cecchin E, Basile A, Bertotti A, Nardini R, Bartoli E.
Renal tubular dysfunction in chronic alcohol abuseeeffects of absti-
nence. N Engl J Med. 1993;329:1927-1934.
33. Bouquegneau A, Dubois BE, Krzesinski JM, Delanaye P. Anorexia
nervosa and the kidney. Am J Kidney Dis. 2012;60:299-307.
34. Hoorn EJ, Zietse R. Disorders of calcium and magnesium balance:
a physiology-based approach. Pediatr Nephrol. 2013;28:1195-1206.
35. Crop MJ, Hoorn EJ, Lindemans J, Zietse R. Hypokalaemia and
subsequent hyperkalaemia in hospitalized patients. Nephrol Dial
Transplant. 2007;22:3471-3477.
- 2014
36. Reddy GS, Jones G, Kooh SW, Fraser D. Inhibition of
25-hydroxyvitamin D3-1-hydroxylase by chronic metabolic acidosis.
Am J Physiol. 1982;243:E265-E271.
37. Wang HF, Tsai SC, Pan MS, Shiao CC. Complete heart block during
potassium therapy in thyrotoxic periodic paralysis. J Emerg Med.
2013;44:61-64.
38. Lin HW, Chau T, Lin CS, Lin SH. Recurring paralysis. BMJ Case Rep.
2009;2009. http://dx.doi.org/10.1136/bcr.07.2008.0577. Epub 2009
Mar 17.
39. Matz R. Risk of hypokalemia developing during the therapy of DKA.
J Diabetes Complications. 1995;9:130-131.
40. Carlotti AP, St George-Hyslop C, Bohn D, Halperin ML. Hypo-
kalemia during treatment of diabetic ketoacidosis: clinical evidence
for an aldosterone-like action of insulin. J Pediatr. 2013;163:
207-212.e1.
41. Huang CL, Kuo E. Mechanism of hypokalemia in magnesium
deciency. J Am Soc Nephrol. 2007;18:2649-2652.
Sung et al Etiology and Management of Hypokalemic Nonperiodic Paralysis 8.e1

Supplementary Table Comparison of Initial Clinical Parame-

ters Between Patients in the Top and Bottom Quartiles of Re-
covery KCl Dosage
Top Quartile Bottom Quartile
Parameters (n 15) (n 15)
Recovery KCl dosage (mmol) 336.0  59.6 161.3  19.2*
Na (135-142 mmol/L) 136.8  7.2 138.4  3.5
K (3.5-5.0 mmol/L) 1.76  0.25 1.98  0.17*
Cl (98-106 mmol/L) 96.1  13.3 102.3  10.6
Bicarbonate (22-26 mmol/L) 30.94  9.95 29.20  7.35
Phosphate (2.6-4.5 mg/dL) 2.78  0.95 3.04  0.79
Total calcium (8.4-10.2 mg/dL) 9.06  1.25 8.57  0.89
Magnesium (1.7-2.55 mg/dL) 1.86  0.41 2.01  0.39
Albumin (3.5-5.7 g/dL) 4.10  0.51 3.74  0.51
Blood urea nitrogen 19.40  6.84 9.80  4.36*
(7-20 mg/dL)
Creatinine (0.7-1.2 mg/dL) 1.16  0.31 0.81  0.19*
Plasma renin activity 23.8  26.3 6.9  8.3*
(0.1-2.0 ng/mL/h)
Plasma aldosterone level 319.9  601.4 267.1  258.6
(12-150 pg/mL)
*Denotes P < .05.