IN· DEPTH REVIEW
Uremic Encephalopathies:
Clinical, Biochemical, and Experimental Features
Cynthia A. Mahoney, M.D. and Allen I. Arieff, M.D.
Patients with renal failure may manifest a variety of neurologic
disorders. Patients with chronic renal failure who have not yet
received dialytic therapy may develop a symptom complex
progressing from mild sensorial clouding to delirium and coma,
with tremor, asterixis, multifocal myoclonus, and seizures.
After the institution of adequate maintenance dialysis therapy,
patients may continue to be afflicted with more subtle nervous
dysfunction, including impaired mentation, generalized weak-
ness, and peripheral neuropathy. These central nervous system
disorders are referred to as uremic encephalopathy. The dialytic
treatment of end-stage renal disease has itself been associated
with the emergence of two distinct, new disorders of the central
nervous system; dialysis dysequilibrium and dialysis dementia.
The dialysis disequilibrium syndrome consists of headache,
nausea, muscle cramps, obtundation, and seizures, and is a
consequence of the initiation of dialysis therapy in some pa-
tients. Dialysis dementia is a progressive, generally fatal en-
A MONG PATIENTS with renal failure, there
have been impressive modifications of both
the duration and quality of life as a result of
dialysis, renal transplantation, and improved med-
ical management. However, patients who have
renal failure continue to manifest a variety of
neurologic disorders. Patients with chronic renal
failure who have not yet received dialytic therapy
may develop a symptom complex progressing from
mild sensorial clouding to delirium and coma, with
tremor, asterixis, multifocal myoclonus, and sei-
zures. Even after the institution of otherwise ade-
quate maintenance dialysis therapy, patients may
continue to be afflicted with more subtle nervous
system dysfunction, including impaired mentation,
generalized weakness, and peripheral neuropathy.
The central nervous system disorders of both un-
treated renal failure and that persisting despite
dialysis are referred to as uremic encephalopathy.
The dialytic treatment of end-stage renal disease
has itself been associated with the emergence of
From Nephrology Research, Veterans Administration Medi-
cal Center and University of California School of Medicine,
San Francisco. Calif.
Supported by Grant No. AM28127-01 from the NIAMDD of
the National Institute of Health. Bethesda. Md.
Reprint requests should be addressed to Allen I. Arieff,
M.D .. Veterans Administration Medical Center. 4150 Clement
Street (J 111), San Francisco, Calif. 94121.
© 1982 by The National Kidney Foundation, Inc.
0272-6386/82/030324-13$02.00/0
324 American Journal of Kidney Diseases, Vol. II, NO.3 (November), 1982
cephalopathy which affects patients on chronic hemodialysis.
There are at least three different forms of dialysis en-
cephalopathy: sporadic, epidemic; and that associated with renal
disease in children. In addition to the foregoing neurologic dis-
eases which are specifically related to uremia and/or dialysis, a
number of other neurologic disorders occur with increased fre-
quency in patients with end-stage renal disease on chronic
hemodialysis. These include subdural hematoma, electrolyte
disorders, vitamin deficiencies, drug intoxication, hypertensive
encephalopathy, and acute trace element intoxication. Renal
transplantation is associated with a variety of central nervous
system infections, reticulum cell sarcoma, and central pontine
myelinosis. The present manuscript will review the clinical,
structural, and biochemical components of those neurologic
disorders which are peculiar to the uremic state and its treatment
with dialysis.
two distinct, new disorders of the central nervous
system: dialysis dysequilibrum and dialysis demen-
tia. The dialysis disequilibrium syndrome consists
of headache, nausea, muscle cramps, obtundation
and seizures, and is a consequence ofthe initiation
of dialysis therapy in some patients. Dialysis de-
mentia is a progressive, generally fatal en-
cephalopathy which affects patients on chronic
hemodialysis. This disease also appears to be a
complication of the therapy for renal failure.
In additon to the foregoing neurologic diseases
which are specifically related to uremia and/or
dialysis, a number of other neurologic disorders
occur with increased frequency in patients with
end-stage renal disease on chronic hemodialysis.
Subdural hematoma is an important consideration
in patients on dialysis who present with headache,
sensorial clouding, or focal neurologic signs. Elec-
trolyte disorders, vitamin deficiencies, drug intox-
ication, hypertensive encephalopathy, and acute
trace element intoxication must be considered with
any altered mental state. Renal transplantation is
associated with a variety of central nervous system
infections, reticulum cell sarcoma, and central
pontine myelinosis.
The present article will review the clinical,
structural, and biochemical components of those
neurologic disorders which are peculiar to the
uremic state and its treatment with dialysis: uremic
encephalopathy, dialysis disequilibrium, and
dialysis dementia. The reader is referred to other
 UREMIC ENCEPHALOPATHIES
reviews for a discussion of those neurologic dis-
orders seen with increased frequency in patients
with renal failure. 1,2
In general, most clinical abnormalities
associated with the uremic syndrome are due,
at least in part, to disordered function of the ner-
vous system. They involve the organism as a
whole, rather than particular organs, and as such
reflect the integrative aspects of the central nervous
system . A fundamental problem facing nephrol-
ogists today is how to assess the adequacy of
dialysis. It has been proposed that the "central
nervous system provides a linkage of relevance be-
tween the chemical lesions, uremic toxins, and ab-
normal metabolic transactions of renal failure on
the one hand, and the clinical, symptomatic lesions
on the other. " 3 Full rehabilitation of patients can-
not be expected without full reversal of the central
nervous system dysfunction.
The central nervous system manifestations of
uremia include a continuum of signs and symptoms
which are characteristic of many metabolic en-
cephalopathies. 2 The earliest symptoms generally
reflect sensorial clouding, and may include loss of
recent memory and impaired ability to concentrate
with depression, delusions , and slurring of speech .
The clinical characteristics of renal failure often
fluctuate from day to day in any given patient, so
that intervals of lucidity may be intermixed with
periods of apathy, confusion or obtundation. Ac-
tion tremor, asterixis , multi focal myoclonus , and
tetany may develop as uremia progresses.
The seizures and coma formerly associated with
advanced renal failure are seldom seen with current
modes of therapy, notably the early institution of
dialysis. The clinical manifestations of acute renal
failure (ARF) do not differ in kind from those seen
in chronic renal failure (CRF), but tend to be more
severe, occur at lower levels of BUN, and in a
more fulminant manner. The neurologic manifesta-
tions of uremia are generally similar whatever the
underlying renal disease, with the exception of sys-
temic diseases which affect the nervous system di-
rectly, such as vasculitis or amyloidosis, and pos-
sibly of patients whose renal failure is due to
diabetes mellitus or primary hyperparathyroidism .
The neurologic manifestations reported with
chronic renal failure have included virtually all ob-
servable clinical abnormalities. 1 Patients with CRF
may develop a sense of sluggishness; easy fatigue
on exertion; insomnia and daytime drowsiness;
325
cutaneous itching; decreased appetite; impaired
ability to focus attention, perform mental arithme-
tic, or express ideas in more than simple language;
erratic memory; slurred speech; vague headaches;
waning sexual interest and potency; restlessness;
emotional irritability and withdrawal; nausea and
vomiting , hypothermia and sensations of coldness.
There may also be paranoid and compulsive per-
sonality changes; bizarre behavior; anxiety; dis-
orientation; confusion; hallucinations; ataxia; tran-
sient mono- , di-, or hemiplegias; aphasia; partial
deafness; and vertigo, all of which may be mixed
with occasional episodes of alertness.
The clinical manifestations of ARF have been
studied in two large series . 4,5 Aberration of mental
status was noted as an early and sensitive index of
neurologic disorder, with lassitude and lethargy
passing rapidly into disorientation and confusion.
Fixed attitudes, torpor, and other signs of toxic
psychosis were common . When the uremia was
untreated and allowed to progress , coma often
supervened. Among these patients, cranial nerve
signs were common, though usually transient.
Both nystagmus and mild facial asymmetries were
common. Visual fields remained normal in most
patients , papilledema was not observed and optic
fundi were normal. Seven of 13 patients had dysar-
thria and difficulty in protruding the tongue past
the lips. Diffuse weakness was always present.
Idiomuscular response to percussion fluctuated
from normal to reduced . Fasciculations were often
present. Marked variation of deep tendon reflexes
was noted in almost all patients, often in an asym-
metrical pattern. Progression to hyperreflexia, with
unsustained clonus at the patella or ankle, was
common .
Electroencephalograms (EEG) have been eval-
uated in over 40 patients with acute renal failure. 4,5
The EEGs of all patients were grossly abnormal
when the diagnosis of renal failure was first estab-
lished . In 16 patients the percentage of EEG power
less than 5 Hz, which is a standard measurement of
the percentage of EEG power devoted to abnormal
(delta) slow wave activity, was over 20 times the
normal value. The EEG was also abnormal with
regard to at least two other parameters of the EEG:
the percentage of EEG frequencies above 9 Hz and
below 5 Hz . These EEG changes were not affected
by dialysis, but returned to normal with recovery
of renal function.
By contrast, the EEG findings in patients with
 326 MAHONEY AND ARIEFF

, =0.63 ,'= 0.40 quencies, loss of the normal alpha rhythm, a shift
80 lOP =1.8 ler), + 6.8 • in both the percentage of EEG frequencies and
P<O.OOl

-r 70

60
N =56
power from higher (> 9 Hz) to lower « 7 Hz)
frequencies and an increase in the burst index. Al-
, ,
~t • •
.::3 50 though these findings are nonspecific, there may
~ also be a myoclonic response to photic stimulation
0
Q.
40
• • at different frequencies, which may be more
'"0
w
30
• specific for uremia. 1-3 Central nervous system dys-
w • •• •
20
••• • function in uremia can also be quantitated by
10 ••• cognition-dependent performance measures. The
0 • ••
• • cognitive functions known to be impaired in
0 4 8 12 16 20 24 28 uremia include sustained attention, selective atten-
Serum creatinine concentration, mg/dl tion, speed of decision making, short-term mem-
ory, and mental manipulation of symbols. 3 The
Fig. 1. Percentage of slow-wave associated (3-7 choice reaction time (CRT) test is probably the
Hz)/(3-13 Hz) EEG power plotted against concurrent
serum creatinine concentrations in 56 azotemic pa- most useful of those employed. It measures sus-
tients with chronic renal failure prior to onset of main- tained attention as well as speed of decision mak-
tenance hemodialysis. (Reproduced with permission ing. Although wide individual differences tend to
of Kidney International. 3 )
obscure the significance of dialysis or transplanta-
CRF are more mild. Several investigations have tion on group results, the improvement in indi-
shown a good correlation between the percentage vidual patients as they go from the azotemic to the
of EEG frequencies and power below 7 Hz and the dialyzed group and from dialysis to transplantation
decline of renal function as estimated by serum is striking (Fig. 2). 3
creatinine (Fig. 1).3 After the initiation of dialysis
there may be an initial period of clinical stabiliza- Fig. 2. Choice reaction
tion during which the EEG deteriorates (up to 6 time (seconds) in five clini-
mo) but it then tends to approach normal values. 3,6 cal groups and in individuals
contributing data to more
However, still more improvement is seen after than one treatment group. In
renal transplantation. 3,7 Patients with CRF demon- the group data, bars repre-
strate a generalized slowing of background fre- sent means (X) and brackets
± SEM of the mean for the
number (N) of the individu-
als in the group. N = nor-
mals: Lo-A = azotemia,
Groups Individuals
serum creatinine < 10 mg/dl;
Hi-A = serum creatinine >
0 .700 0.700 10 mg/dl; D = during stable
maintenance hemodialysis;
T = following successful
0.600 0.600 renal transplantation. For
individuals contributing data
to more than one treatment
situation (A,D, or T), each
0.500
data point represents the
~
1-' single value or the mean of 2
a: to 25 separate values from
u
0.400 each patient in each treat-
ment situation. Open circles
(0) identify four patients
who traversed and were
studied in all three clinical
situations: A (Hi-A), D, and T.
Levels of statistical sig-
nificance (P) are shown for
the indicated comparisons:
P N NS NS = not significant; (*) = P
Lo - A NS NS NS < 0.05; (**) = P < 0.01. (Re-
HI - A
produced with permission of
o NS P
Kidney International. 3)
 UREMIC ENCEPHALOPATHIES
Biochemical Changes in Brain
To determine possible causes of the EEG abnor-
malities and clinical symptoms observed in pa-
tients with either acute or chronic renal failure,
biochemical studies have been carried out in brain
of both patients and laboratory animals. Most
studies have been done in animals with ARF.
Measurements have included brain intracellular pH
and concentrations of Na+, K+, Cl-, fhO, AI+3,
Ca+ 2, Mg+2, urea, adenine nucleotides (creatine
phosphate, ATP, ADP, AMP), lactate, and (Na+
+ K+)-activated adenosine triphosphatase (ATP-
ase). In patients with acute renal failure, content of
fhO, K+ and Mg+2 is normal, while Na+ is mod-
estly decreased and Al +3 slightly elevated. How-
ever, cerebral cortex Ca +2 content is almost twice
the normal value. 1,5 Similar findings have been ob-
served in dogs with acute renal failure. 8 Permeabil-
ity of uremic brain (in rats) to inert molecules, such
as inulin and sucrose, is increased. Entry into brain
of Na+ is impaired, while that of K+ is increased.
The permeability of brain to weak acids, such as
sulfate, penicillin, and dimethadione, is normal to
low. 9 - metabolic rate, along with elevated glucose and
Alterations of cerebral metabolism that might be
related to the aforementioned changes in permea-
bility have also been studied in animals. In brain of
rats with acute renal failure, van den Noort and
co-workers found that creatine phosphate, ATP,
and glucose were increased, but there were corres-
ponding decreases in creatine, AMP, ADP, and
pH W (nMI
7.4
1.3
o normal
1.2 uremia
1.1
7.0
6.9
6..8
01.
BLOOD CSF BRAIN pHi MUSCLE pHi
Fig. 3. The pH of arterial blood and cerebrospinal
fluid (CSF), and the intracellular pH (pHi) in brain
(cerebral cortex) and skeletal muscle. In animals with
uremia (serum creatinine, 11.8 ± 0.4 mg/dl), there was
a significant fall in arterial pH. However, there was no
change in pH of CSF, or of the pHi in brain or muscle.
(Reproduced with permiSSion of The American Society
for Clinical Investigation.' 2)
327
INTRACELLUL AR pH IN UREMIC PATIENTS
160 r
pH
pH ,
140
130
1.20
7.10
1.00
6 .90
6 .80
610
660
o
BLOOD MUSCLE WHOLE BOOY WHITE
BLOOD
CELLS
Fig. 4. Arterial pH in patients compared with in-
tracellular pH (pHi) in skeletal muscle, "whole body,"
and white blood cells. Although there is a significant
extracellular acidemia, pHi is normal in muscle,
"whole body," and white blood cells. (Reproduced
with permission of W. B. Saunders Co.')
lactate. Total brain adenine nucleotide content and
(Na+ + K+) activated ATPase were normal to low.
The uremic brain utilized less ATP and thus failed
to produce ADP, AMP, and lactate at normal rates.
There was a corresponding decrease in the brain's
low lactate levels. 10 Other studies of uremic brain
have shown a decrease in cerebral oxygen con-
sumption. 1 Patients with chronic renal insuf-
ficiency (GFR < 20 ml/min) have decreased brain
uptake· of glutamine and increased ammonia up-
take. The relevance of these findings, in terms of
neurotransmitters or other brain function, is un-
known.
In animals with either acute or chronic renal
failure, both urea concentration and osmolality are
50 similar in brain, CSF, and plasma. The solute con-
tent of brain in animals with acute renal failure is
63
such that essentially all of the increase in brain
80 osmolality is due to an increase of brain urea con-
100
centration. However, in animals with chronic renal
failure, about half of the increase in brain osmolal-
126 ity is due to the presence of undetermined solute
156
(idiogenic osmoles). 1
In animals who have acute renal failure and
metabolic acidosis, the intracellular pH (pHi) of
brain, and skeletal muscle as well, is normal (Fig.
3). In patients with renal failure, intracellular pH
has been reported to be normal in both skeletal
muscle and leukocytes, as well as in the "whole
body" (Fig. 4).1 The pH of CSF has also been
shown to be normal in both patients and laboratory
animals with renal failure. Thus, despite the pres-
328
ence of extracellular metabolic acidemia, brain pHi
is normal. 12
Pathology
Pathologic studies of brain have been reported in
over 400 patients dying with chronic renal failure. 1
Subdural hemorrhages used to be very common in
dialyzed subjects, being reported in about
1% - 3% of autopsied uremic patients prior to
1974. In addition, intracerebral hemorrhages were
reported to be present in about 6% of dialysis pa-
tients who expired, but these data are colored by
the fact that many patients were receiving oral an-
ticoagulants. Histologic changes in brain are prob-
ably nonspecific. Cerebral edema is not present in
brain of patients or laboratory animals with chronic
renal failure, either by biochemical or histologic
criteria. Generalized but variable neuronal degen-
eration is often present but is inconsistent as to
location. Among patients dying of uremia per se,
there is some necrosis of the granular layer of the
cerebral cortex. Small intracerebral hemorrhages
and necrotic foci are seen in about 10% of uremic
patients. Focal glial proliferation is found in about
2% of uremic patients. In general, changes in brain
of patients who died with CRF are probably
nonspecific and related more to any concomitant
underlying disease state.
Pathophysiology - Role of PTH
Although there are a multitude of factors that
contribute to uremic encephalopathy, many inves-
tigators have shown no correlation between en-
cephalopahy and any of the commonly measured
indicators of renal failure (BUN, creatinine, bicar-
bonate, pH, potassium). 2In recent years, there has
been considerable discussion of the possible role of
parathyroid hormone (PTH) as a uremic toxin. In
particular, there is a substantial amount of evi-
dence which suggests that there are toxic effects of
PTH on the central nervous system (CNS). The
evidence for neurotoxicity of PTH in both humans
and laboratory animals will now be reviewed.
In patients dying with acute or chronic renal
failure, the calcium content in brain cerebral cortex
is significantly elevated. 5.13 Dogs with acute or
chronic renal failure show increases of brain gray
matter calcium and EEG changes similar to those
seen in humans with renal failure. 8 In dogs, both
the EEG and brain calcium abnormalities can be
prevented by parathyroidectomy. Conversely, they
MAHONEY AND ARIEFF
can be reproduced by administration of parathyroid
hormone to normal animals, but not by elevation of
the (calcium) (phosphate) product nor by hypercal-
cemia induced by vitamin D intoxication. Thus,
PTH appears essential to produce central nervous
system manifestations in the dog model of uremia.
PTH is known to have CNS effects in humans
even with normal renal function. Neuropsychiatric
symptoms are becoming the most frequent man-
ifestation of primary hyperparathyroidism. 14 Pa-
tients with primary hyperparathyroidism also have
EEG changes similar to those observed in patients
with acute renal failure. 5,13 The common de-
nominator appears to be elevated parathyroid hor-
mone. In patients with acute renal failure the EEG
is abnormal within 18 hr of the onset of renal fail-
ure and is generally not affected by dialysis for
periods of up to 8 wk. 5 During this interval, pa-
tients with acute renal failure have been shown to
have elevated levels of PTH in plasma. In patients
with either primary or secondary hyper-
parathyroidism, parathyroidectomy results in an
improvement of both EEG and psychologic test-
ing, suggesting a direct effect of PTH on the cen-'
tral nervous system. 13 Similarly, dialysis results in
a decrement of brain (cerebral cortex) calcium
toward normal in both patients and laboratory ani-
mals with renal failure, concomitant with im-
provement of the EEG. 1.15 In uremic patients, both
EEG changes and psychologic abnormalities are
improved by parathyroidectomy or medical sup-
pression of PTH. 13 ,16 Thus, PTH and/or a high
brain calcium content are probably responsible, at
least in part, for many of the encephalopathic man-
ifestations of renal failure.
The mechanism by which PTH might disturb
central nervous system function is unclear, al-
though the occurrence of psychosis in patients with
hyperparathyroidism has been well doc-
umented. 1,14 The increased calcium content in such
diverse tissues as skin, cornea, blood vessels,
brain, and heart in patients with hyper-
parathyroidism suggests that PTH may somehow
facilitate the entry of Ca +2 into such tissues. The
finding of increased calcium in the brain of both
dogs and patients with chronic renal disease and
secondary hyperparathyroidism is consistent with
the conception that part of the central nervous sys-
tem dysfunction and EEG abnormalities found in
acute or chronic renal failure may be due to a PTH
mediated increase in brain calcium. Calcium is es-
UREMIC ENCEPHALOPATHIES
sential for the function of a large number of in-
tracellular enzyme systems, and an increased brain
calcium content could disrupt cerebral function by
interfering with any of them. It is also possible that
PTH itself may have a detrimental effect on the
central nervous system .
In summary, of the long list of proposed uremic
toxins , none has been shown to correlate better
than creatinine clearance with any manifestations
of uremia . 1 A pathophysiologic role in the central
nervous system dysfunction of uremia has been
clearly established thus far only for parathyroid
hormone. Pathologic and histologic examinations
have generally shown minor, nonspecific abnor-
malities. However, there is a notable lack of ultra-
structural investigations of the central nervous sys-
tem in uremia. Neither brain edema nor in-
tracellular acidosis occurs . The functional sig-
nificance of the altered brain permeability, de-
creased brain sodium , and presence of idiogenic
osmoles is uncertain. The low turnover rate of high
energy phosphates seen in rats with acute renal
failure and the decreased cerebral oxygen con-
sumption may as easily be a result as a cause of the
uremic state. The role of neurotransmitters and
other peptide hormones in uremia are additional
areas which have not been addressed.
DIALYSIS DISEQUILIBRIUM
Dialysis disequilibrium syndrome (DDS) is a
clinical syndrome which occurs in patients being
treated with hemodialysis (HD). The syndrome has
been described since about 1962 17 and may include
headache, nausea, emesis, blurring of vision, mus-
cular twitching, disorientation, hypertension,
tremors, and seizures. The electroencephalogram
(EEG) may be abnormal, 17 but this has been dis-
puted. 7 More recently , the syndrome of DDS has
been expanded to include milder symptoms , such
as muscle cramps, anorexia , restlessness , and diz-
ziness. 18
The clinical manifestations of DDS have been
summarized.1,15,17 For convenience, they
have been divided into major and minor symp-
toms. Mild forms of DDS may be manifested by no
more than restlessness and severe headache, which
may occur during or soon after hemodialysis . This
is commonly followed by nausea and vomiting,
often accompanied by blood pressure elevation .
These symptoms may be accompanied by disorien-
tation and tremors . Seizures and cardiac ar-
329
rhythmias have been reported but are uncommon.
The seizures have usually been of the grand mal
type , although focal tremors have been observed.
The symptoms are usually self-limited but recov-
ery may take several days. In rare instances the
seizures may lead to coma. It appears that modem
methods of dialysis have altered the clinical picture
of DDS . Most reports of seizures, coma, and death
were reported prior to 1970. The symptoms of
DDS as reported in the last decade (1972-1982)
have generally been mild , consisting of nausea,
weakness, headache, fatigue, and muscle cramps.
It is also unclear whether any patient ever actually
died from DDS per se, or from other associated
neurologic complications of dialysis , such as acute
stroke, subdural hematoma, subarachnoid hemor-
rhage , or hyponatremia.
There has been much debate about the
pathogenesis of DDS. In early studies on patients
treated with HD, a consistent el~vation in cerebro-
spinal fluid (CSF) pressure was observed. 17 In a
few patients who died with , but probably not be-
cause of DDS, the presence of brain edema was
demonstrated. 1 Because of the aforementioned, it
has been suggested that brain edema is a major
factor underlying the pathogens is of DDS. Early
clinical investigation in patients with DDS in-
cluded evaluation of pressure and composition of
the CSF. In patients with DDS, findings included a
persistent elevation of CSF pressure, and levels of
urea which were higher in CSF than in blood. 17
These findings led to the early formulation that
cerebral edema was responsible for most of the
manifestations of DDS.
The presence of brain edema in some patients
with DDS, along with the observation that in pa-
tients undergoing rapid HD, urea levels are higher
in CSF than in blood, led to formulation of the
"reverse urea " hypothesis . Simply put , the "re-
verse urea" effect states that steady-state concen-
trations of urea are similar in plasma, CSF and
brain tissue water. With rapid HD, it is further
assumed that clearance of urea will be more rapid
from plasma than from brain. This assumption
stems from the observed slower clearance of urea
from CSF than from plasma 3 in patients undergo-
ing rapid HD. The decrement of plasma urea leads
to a fall in plasma osmolality . It was postulated
that, with a slower clearance of urea from brain,
osmolality of brain tissue remained elevated above
that of plasma, leading to a net movement of water
330 MAHONEY AND ARIEFF

•
pHi H+ (nM I
from plasma into brain, with resultant cerebral uremic
edema. However, the "reverse urea" hypothesis l2 80
has not been supported by experimental investiga- II slow HD

tions. The rate of urea clearance from brain essen-
tially parallels that of plasma during rapid
hemodialysis in uremic dogs. 15
There is an alternate explanation for the cerebral
edema which does have experimental support.
Along with the delayed clearance of urea from the
CSF, there is evidence for a paradoxical acidemia
of CSF in both patients and laboratory animals
with DDS. In some patients treated with rapid
hemodialysis, there is a fall in pH of the CSF de-
spite a rise in arterial pH. 1 A similar fall in pH of
CSF occurs after rapid hemodialysis of experi-
mental animals with acute renal failure (Fig. 5).12
The decrement of pH in CSF may be a major factor
in the pathogenesis of DDS. Along with the dec-
rement in pH of CSF, there is an observed decrease
in the pHi of brain. In dogs with acute renal failure
(ARP), pHi is normal in brain and skeletal muscle
(Fig. 6). The increased H+ ion activity in brain is
the pHi of muscle, but a significant decrement is
observed in the pHi of cerebral cortical gray matter
(Fig. 6). The increased H+ ion activity in brain is
accompanied by an increase of brain osmole con-
tent, which secondarily results in an increase of
brain water. The cerebral edema which thus occurs
is probably the cause of the observed clinical man-
ifestations of DDS. The presence of postdialysis
brain swelling has recently been documented by
BLOOD CSF
belore HD alterHO belo.. HD alter HD
l5
l4
0
7.1
rapid HD
100
7.0
6.9
6.B 158
ot BRAIN MUSCLE
Fig. 6. The intracellular pH (pHi) of brain (cerebral
cortex) and skeletal muscle in uremic dogs, and dogs
treated with either rapid or slow hemodialysis (HD).
After slow HD, there is a small but not significant fall in
brain pHi. After rapid HD, there is a highly significant
fall in brain pHi. The pHi of muscle is unaffected by
either uremia or HD. (Reproduced with permission of
The American Society for Clinical Investigation. 12)
CAT scan in patients with chronic renal failure.
The source of the increased brain H+ ion is not
known.
Patients who have renal failure and are being
treated with HD may manifest the symptoms of
headache. nausea, emesis, muscle cramps, and
seizure activity because of causes other than DDS.
A differential diagnosis of such patients is shown
in Table 1. Recently, the diagnosis of DDS has
become a "wastebasket" for a number of disorders
which can occur in patients with renal failure and
may affect the central nervous system. It should be
stressed that the diagnosis of DDS should be one of
exclusion.
32 Table 1. Disorders Which May Mimic Dialysis
Disequilibrium Syndrome
40

~ ~
1. Copper intoxication
2. Subdural hematoma
l3 50
3. Uremia, per se
4. Nonketotic hyperosmolar coma with hyperglycemia
12 63
5. Cerebral embolus secondary to shunt declotting
6. Acute cerebrovascular accident
l1 80
pH H+ (nM) 7. Dialysis dementia
8. Cardiac arrhythmia
Fig. 5. Changes in arterial blood and cerebrospi- 9. Depletion syndrome
nal fluid (CSF) during rapid hemodialysis (HD). During
10. Malfunction of fluid proportioning system
HD, the blood pH increases from 7.22 ± 0.02 to 7.31 ±
0.01, while bicarbonate rises from 13.7 ± 1.0 to 17.7 ± 11. Excessive ultrafiltration with hypotension and seizures
0.7 mmol/liter. The simultaneously determined pH in 12. Hypoglycemia
CSF falls from 7.22 ± 0.02 to 7.31 ± 0.01, while bicar- 13. Wernicke's encephalopathy
bonate falls from 18.9 ± 1.1 to 15.4 ± 0.5 mmol/liter. 14. Hyperparathyroidism with hypercalcemia (Serum Ca
There was no change in arterial blood PC02. Heavy above 14 mg/dl)
lines are mean ± SE, while each thin line represents a 15. Malignant hypertension
single dog study before and after rapid HD. (Repro- 16. Hyponatremia (serum Na below 125 mEq/liter)
duced with permission of The American Society for 17. Nickel intoxication
Clinical Investigation. 12)
UREMIC ENCEPHALOPATHIES
With this understanding of the pathophysiology
of the disequilibrium syndrome, management can
be divided into preventative modalities and thera-
peutic maneuvers. In several studies, this syn-
drome has been treated by addition of osmotically
active solute (glucose, glycerol, albumin, urea,
fructose, NaCl, mannitol) to the dialysate, or by
substitution of sodium bicarbonate for sodium lac-
tate (or acetate) in the dialysate. 1,17,18 The purpose
of such maneuvers has been to minimize rapid
alterations in plasma osmolality or bicarbonate dur-
ing dialysis. In uremic animals undergoing rapid
hemodialysis, addition of glycerol or mannitol to
dialysate prevents many of the manifestations of
experimental DDS. 1 Studies on uremic patients
undergoing hemodialysis with either glycerol,
NaCl, or mannitol added to dialysate suggest that
these agents often improve symptoms which may
be associated with DDS. 1,18 However, the
aforementioned studies, as well as most others
where osmotically active solute was added to the
dialysate, were carried out on stable chronic
dialysis patients. Additional studies are needed in
patients with acute renal failure, or in those with
chronic renal failure undergoing their initial few
hemodialyses. Evaluation of the efficacy or pre-
vention of DDS by addition of solutes to the dialy-
sate must await these additional investigations.
The occurrence of DDS can probably be best pre-
vented by employing short « 4 hr), frequent
(every 1-2 days) dialysis at low blood flow rates
« 200 ml/min) in those patients most likely to
develop DDS. 1 Pure ultrafiltration may be of ben-
efit in patients who require fluid removal in excess
of their dialytic requirements. The role of ultrafilt-
ration followed by dialysis in the prevention of
DDS is being evaluated. Peritoneal dialysis should
also be considered in patients at great risk of DDS.
DIALYSIS DEMENTIA
Dialysis dementia (also called dialysis en-
cephalopathy) is a progressive, frequently fatal
neurologic disease which was first described in the
early 1970s. The disease is seen almost exclusively
in patients on chronic hemodialysis. Although the
early literature focused on the distinctive
neurologic findings, 19 more recent reports from
both Europe and the United States have suggested
that some forms of dialysis dementia may be part
of a multisystem disease which includes en-
cephalopathy, osteomalacic bone disease, proxi-
331
mal myopathy, and anemia. 20-24 In addition, sev-
eral series have noted nonspecific gastrointestinal
complaints in almost half the patients.
The first report from Denver in 1972 19 was soon
followed by reports throughout the world. 1 Most
affected patients had been on chronic hemodialysis
for over 2 yr before the onset of symptoms. Early
manifestations consisted of a mixed dysarthria-
apraxia of speech with slurring, stuttering, and
hesitancy. Personality changes, including psycho-
sis, led to dementia, myoclonus, and seizures.
Symptoms might initially be intermittent and were
often worse during dialysis, but generally became
constant and progressed to death within 6 mo.25
The disease was devastating to those units af-
fected, which reported incidences up to 40%.26
Speech disturbances occur in 90% of patients, af-
fective disorders culminating in dementia in 80%,
motor disturbances in 75%, and convulsions in
60% - 90%. In contrast to this fairly distinct clini-
cal picture, brain histology has been normal or
nonspecific.
Early in the disease, the EEG shows multi focal
bursts of high amplitude delta activity « 4/sec
waves) with spikes and sharp waves, intermixed
with runs of more normal appearing background
activity. These EEG abnormalities may precede
overt clinical symptoms by 6 mo. As the disease
progresses, the normal background activity also
deteriorates to slow frequencies. 23,25 The EEG has
been said to be pathognomonic, but may also be
seen in other metabolic encephalopathies, such as
hypophosphatemia and hypercalcemia. The diag-
nosis depends on the presence of the typical clini-
cal picture, confirmed by the characteristic EEG
pattern.
The etiology of this syndrome remains contro-
versial. Although an increase in brain aluminum
content has been strongly implicated in some cases
of dialysis dementia, the evidence is less convinc-
ing in others. Dialysis dementia most likely repre-
sents a syndrome complex which is the final com-
mon pathway for a variety of etiologic agents. At
this stage of our knowledge it seems useful to sub-
divide dialysis dementia into three categories: (1)
an epidemic form which is related to contamination
of the dialysate, probably with aluminum; (2)
sporadic cases in which aluminum intoxication is
less likely to be a contributory factor; and (3) de-
mentia associated with congenital or early child-
hood renal disease. This entity has been reported in
332
several children who were never dialyzed or ex-
posed to aluminum compounds. It now appears
that these early childhood cases represent devel-
opmental neurologic defects resulting from expo-
sure of the growing brain to a uremic envi-
ronment. 27
Aluminum intoxication was first implicated in
this disorder by Alfrey and colleagues in 1976. 28
Aluminum content of brain gray matter was ele-
vated II-fold in their patients with dialysis demen-
tia and 3-fold in their patients on chronic
hemodialysis without dialysis dementia. relative to
normal controls. Bone and other soft tissue bur-
dens of aluminum were also increased. No consist-
ent differences were found in a wide variety of
other trace elements studied. Oral phosphate bind-
ers containing aluminum were originally suspected
as the source of the aluminum.
Significant absorption of oral aluminum can
occur in patients with chronic renal failure, and
may account for some of the sporadic cases occur-
ring in undialyzed patients and areas without high
dialysate aluminum. 28,29 However. the weight of
evidence is against oral aluminum as the major
source. The use of such compounds has been al-
1-2
}-S
6-10
CASES
CASES
CASES
..•
•
MAHONEY AND ARIEFF
most universal, yet the vast majority of cases have
occurred in an almost epidemic form with striking
geographic clusters (Fig. 7). Instead, epidemio-
logic studies have shown a strong association
between aluminum in the water used to prepare
dialysate and both dialysis encephalopathy and
fracturing dialysis osteodystrophy. High water
aluminum levels most often arise from the use of
aluminum salts to precipitate organic debris and
clarify municipal water supplies. Seasonal varia-
tions in the use of aluminum may result in levels
ranging from 1. 0 to over 1600 /Lg/liter. Water soft-
ening does not remove aluminum effectively. 21
Although reverse osmosis does remove up to 80%,
significant residual levels may remain if the initial
levels are high. Deionization (usually following
reverse osmosis) is effective in decreasing alumi-
num to levels of < I /Lg/liter. Thus, the aluminum
content of dialysate depends not only on the en-
dogenous level, the municipal use of aluminum
and the season, but also on the method of water
treatment used to prepare dialysate.
In a survey by the European Dialysis and Trans-
plant Association, 92% of patients with dialysis
encephalpathy had used untreated or softened wa-
Fig. 7. Geographical dis-
tribution of 150 cases of
dialysis dementia in Europe,
1976-77. Centers which re-
ported numerous cases are
shown with larger triangles.
Note that many areas of high
population density, e.g.,
Thames valley, the Ruhr,
and Italy, did not report
many cases despite large
numbers of patients on
treatment. (Reproduced with
permiSSion of Lancet. 22 )
UREMIC ENCEPHALOPATHIES
ter, as opposed to 6% who had used deionized
water. The use of the two forms of water treatment
was equally distributed in the same population. 26
These findings may largely explain the increase of
dialysis dementia in home dialysis patients, who
are less likely to use deionized water. Outbreaks of
dialysis encephalopathy have been associated with
the onset of the use of aluminum for municipal
water purification 25 and with dysfunction of the
deionizer. 30
The strongest evidence that aluminum may be
directly responsible for dialysis dementia comes
from an outbreak in Eindhoven, Holland. 20 Two
dialysis units there used the same untreated munic-
ipal water supply (60 JLg/liter aluminum) and the
same dialysate concentrate. One unit had six cases
of dialysis dementia in a 9-mo period, yet the sister
unit was unaffected. A corroding aluminum anode
in the water heating system of the affected unit was
found to be adding aluminum to the dialysate,
elevating the dialysate aluminum to over 800 JLg/
liter. Lead, copper, and mercury levels in the
dialysate were normal. When the unit was moved
to a new hospital, no new patients developed en-
cephalopathy or fractures. The possibility of other
contaminating substances in the plumbing system
seems unlikely, but cannot be entirely excluded. In
the epidemiologic study reported by Parkinson, pa-
tients using deionized water were excluded, yet a
strong linear correlation was still seen between
dialysis dementia and water aluminum concentra-
tion over 50 JLg/liter (r = 0.69). 26 It is unlikely
that aluminum was simply a common marker for
another toxic agent in all the locales surveyed. The
incidence of dialysis dementia does not correlate
with length of time on dialysis, but does correlate
with the duration of exposure to high aluminum
dialysate. 21.30
Even low levels of aluminum in the dialysate
will result in a net retention of aluminum in the
body. Aluminum transferred across the dialyzer
binds to a nondialyzable plasma constituent, result-
ing in higher levels in plasma than in the dialysate.
It has been estimated that a dialysate aluminum
concentration of only 70 JLg/liter results in a mini-
mum positive aluminum balance of 300 JLg/day. 31
Significant amounts of aluminum are also absorbed
from the gastrointestinal tract after oral administra-
tion of aluminum hydroxide. Normal individuals
can absorb up to 1 mg/day, while patients with
chrontc renal failure may absorb even more. Since
333
aluminum is normally excreted by the kidney, this
can lead to significant aluminum retention in pa-
tients with renal insufficiency.
Excess aluminum accumulates in brain, bone,
and other soft tissues. The correlation between
serum levels and tissue levels is poor. Either
changing to a low aluminum dialysate or renal
transplantation may result in relatively normal
plasma aluminum levels despite persistently high
tissue levels. 30 Since serum levels generally reflect
dialysate aluminum levels, most authors have not
found them useful in separating patients with
dialysis dementia from other patients dialyzed in
the same center. No one disputes that brain alumi-
num is elevated in dialysis dementia, but it has
been questioned whether high brain aluminum is
the cause of neurotoxicity rather than an associa-
tion. Although Alfrey and McDermott were able to
show a good separation between brain cortical gray
matter aluminum in hemodialyzed patients with
and without dialysis encephalopathy, other authors
have noted considerable overlap. 1,28,30,32 Surpris-
ingly, brain aluminum levels are higher in patients
with chronic renal failure not yet on dialysis than
their dialyzed counterparts. Elevated brain alumi-
num levels are also seen in patients with hepatic
encephalopathy and metastatic cancer, though they
are generally lower than in patients with dialysis
encephalopathy (Fig. 8). 1 Unfortunately, most
studies have reported brain aluminum levels in
normals rather than the appropriate control-patients
with chronic renal failure with and without
hemodialysis. There is no obvious threshold level
for brain aluminum which determines toxicity. In
general though, levels above 15 JLg/kg dry weight
are associated with the encephalopathy, while
levels below 5 JLglkg are not.
Deionization of the water used to prepare dialy-
sate has been the most consistently effective pre-
ventive measure. 20,21,25,30 However, deionization
may be beneficial by removing any number of
other agents. Other trace elements may be present
in water which can result in central nervous system
toxicity. Such elements include cadmium, mer-
cury, lead, manganese, copper, nickel, thallium,
boron, and tin. 33 Among these potentially
neurotoxic elements no one has measured brain
content of cadmium, mercury, nickel, thallium,
vanadium, or boron. Manganese has been found to
be increased in cortical white matter in the eight
encephalopathic patients in whom it was meas-
334
lloi
I dlalYll1 dementia
Per Cenl 01 CONTROL
700
600
Fig. 8. Brain (gray mat-
ter) aluminum (AI+3) content 500
in normal patients and those
who had either hepatic coma, 400
Alzheimers, age > 70,
dialysis dementia, renal 300 a.e ) 70 yr
failure, or metastatic cancer.
The brain AI+3 is significantly 200
above normal in all of these
groups. (Reproduced with 100
permission of W. B.
Saunders.1)
ured. 34 These patients also had elevated aluminum
levels in gray matter. Copper, zinc, and iron levels
were normal. It is important to realize that tin was
the first metal implicated in the etiology of dialysis
dementia, before aluminum had been routinely
measured. 28 Thus, still another agent which is not
currently measured may prove to correlate with
dementia better than aluminum, especially in the
sporadic cases. The failure of most reports to in-
clude results of hemodialyzed controls also makes
attribution of toxicity difficult. The dialysate itself
must be measured if trace element toxicity is sus-
pected. A number of metals may potentially leach
into the dialysate water from the plumbing system
after the deionizer, including aluminum, nickel,
cadmium, tin, and manganese. Other etiologies
have been proposed for dialysis dementia. 1 Most
have not been directly substantiated, although a
slow virus has been isolated from the brain of one
patient.
Most of the controversy over the etiology of
dialysis dementia has involved those cases which
occur sporadically. As noted previously, dialysate
aluminum levels are not always elevated, the use
of aluminum phosphate binders is no different than
in unaffected patients, and brain aluminum levels
overlap those of unaffected patients. There are
very few studies on the etiology of these sporadic
cases. Arieff found that brain aluminums were ele-
vated in a small group of encephalopathic patients,
but values overlapped those of unaffected pa-
tients.1
The largest group of . 'sporadic cases" has been
MAHONEY AND ARIEFF
BRAIN ALUMINUM
CRF
ARF
hepatic metastatic
coma canc.r
Alzh.lmerl
reported from Nashville. 29 The incidence of
dialysis dementia in the area is 5%, despite the use
of deionized water for dialysate with aluminum
levels of < 5 /-Lg/liter. Bone disease was not clini-
cally apparent in this group. Serum aluminum
levels in the encephalopathic group were 3 - 4
times higher than other dialyzed patients, despite
equivalent prescribed doses of aluminum contain-
ing phosphate binders. 35 These results suggest
greater absorption and/or retention of aluminum in
this group of encephalopathic patients. Anecdotal
reports of improvement in several patients after
withdrawal of oral aluminum require further
confirmation. No other metals were measured in
the Nashville study.
Despite these unresolved questions, most out-
breaks of dialysis dementia have been associated
with high levels of aluminum in the dialysate.
Lowering the dialysate aluminum to levels below
20 /-Lg/liter, usually by deionization, prevents the
onset of the disease in patients newly begun on
dialysis. New cases may continue to appear in
those patients who were previously exposed to the
high aluminum dialysate, although the course is
milder and mortality is decreased to 25%.30 In pa-
tients with overt disease, eliminating the source of
aluminum has resulted in improvement in some 29 ,30
but not all patients. 22 Transplantation has generally
not been helpful in patients with advanced disease.
Diazepam or clonazepam are useful early in the
disease, but become ineffective later on and do not
alter the final outcome. 19,30
Treatment of sporadic cases, in which the etiol-
 UREMIC ENCEPHALOPATHIES
Table 2. Differential Diagnosis of Dialysis Dementia
Metabolic Encephalopathies
Hypercalcemia
Hypophosphatemia
Hypoglycemia
Hyperosmolarity
Hyponatremia
Overt uremia
Drug intoxications
Trace element intoxications: manganese, mercury, lead,
nickel, thallium, boron, vanadium, chromium, tin,
cadmium
Hyperparathyroidism
Hypertensive Encephalopathy
Dialysis Disequilibrium
Structural Lesions
Subdural hematoma
Normal pressure hydrocephalus
Stroke
ogy is not clear, is of course more difficult. Every
effort should be made to identify a treatable cause.
Dialysis dementia must be differentiated from
other metabolic encephalopathies, such as hyper-
calcemia and hypophosphatemia, hyper-
parathyroidism, acute heavy metal intoxication,
drug intoxications (notably the long acting seda-
tives) and structural neurologic lesions, such as
subdural hematoma (Table 2). Because of the low
incidence, the uncertain etiology, and the poor cor-
relation of plasma with tissue aluminum levels,
screening tests have not generally been employed.
In the European experience, a high incidence of
bone disease occurred after a shorter exposure to
high aluminum dialysate than required to produce
encephalopathy. Thus, screening EEGs and
neurologic exam with emphasis on speech disturb-
ances might be useful in the early detection of en-
REFERENCES
1. Arieff AI: Neurological complications of uremia. in
Brenner BM. Rector FC, Jr (eds): The Kidney. Philadelphia.
W. B. Saunders Co., (Chapter 44) 1981, pp 2306-2343
2. Raskin NH, Fishman RA: Neurologic disorders in renal
failure. N Engl J Med 294:143-148,1976
3. Teschan PE. Ginn HE, Bourne JR, et aI: Quantitative
indices of clinic uremia. Kidney Int 15:676-697, 1979
4. Locke S, Merrill JP, Tyler HR: Neurologic complications
of acute uremia. Arch Intern Med 108:75-86. 1961
5. Cooper JD. Lazarowitz VC, Arieff AI: Neurodiagnostic
abnormalities in patients with acute renal failure: Evidence for
neurotoxicity of parathyroid hormone. J Clin Invest 61: 1448-
1455. 1978
6. Kiley J, Hines 0: Electroencephalographic evaluation of
uremia. Arch Intern Med 116:67-73, 1965
335
cephalopathy in patients with the typical os-
teomalacic bone disease.
There is not sufficient clinical evidence to war-
rant abandoning the general use of oral aluminum
gels as phosphate binders. No equally effective
substitute is available. Patients would risk severe
hyperparathyroidism, a disease of known etiology
and high incidence, in an attempt to prevent a dis-
ease of low incidence which may not be related to
oral aluminum. In contrast, the evidence is quite
strong that high aluminum dialysate is associated
with a substantial risk of encephalopathy and os-
teodystrophy. Water should be purified so that
dialysate contains no more than 20 J.Lg/liter alumi-
num, and preferably less than 1 J.Lglliter. Deioniza-
tion of all water supplies is preferable. Measure-
ment of dialysate aluminum levels is mandatory if
cases of encephalopathy or osteomalacia are sus-
pected.
One of the major unresolved questions in the
field is what triggers the onset of dementia in pa-
tients at risk. The onset of dementia has been asso-
ciated with hospitalization for another medical ill-
ness in 70% of cases. 30 One theory is that im-
mobilization causes aluminum release from bone,
acutely increasing serum levels. Parathyroid hor-
mone has been shown to enhance gut absorption of
aluminum and to retard mobilization from tissues,
including brain, in experimental animals. 36 Al-
though most patients with chronic renal failure
have elevated PTH levels, the levels actually tend
to be suppressed in patients with osteomalacic os-
teodystrophy. There is also preliminary in vitro
evidence that aluminum suppresses PTH secretion.
Thus, the possible role of PTH in dialysis en-
cephalopathy remains unclear.
7. Kiley JE, Woodruff MW, Pratt KL: Evaluation of en-
cephalopathy by EEG frequency analysis in chronic dialysis
patients. Clin Nephrol 5:245-250, 1976
8. Guisado R, Arieff AI. Massry SG, et aI: Changes in the
electroencephalogram in acute uremia: Effects of parathyroid
hormone and brain electrolytes. J Clin Invest 55:738-745,
1975
9. Fishman RA: Permeability changes in experimental
uremic encephalopathy. Arch Intern Med 126:835- 837, 1970
10. van den Noort S, Eckel RE, Brine K. et aI: Brain
metabolism in uremic and adenosine-infused rats. J Clin Invest
47:2133-2142. 1968
11. Deferrari G. Garibotto G, Robaudo C. et al: Brain
metabolism of amino acids and ammonia in patients with
chronic renal insufficiency. Kidney Int 20:505- 510. 1981
336
12. Arieff AI. Guisado R. Massry SG. et al: Central nervous
system pH in uremia and the effects of hemodialysis. J Clin
Invest 58: 306- 311, 1976
13. Cogan M. Covey C. Arieff AI, et al: Central nervous
system manifestations of hyperparathyroidism. Am J Med
65:963-970, 1978
14. Heath H. Hodgson SF, Kennedy MA: Primary hyper-
parathyroidism. N Engl J Med 302:189-193. 1980
IS. Arieff AI, Massry SG, Barrientos A, et al: Brain water
and electrolyte metabolism in uremia: Effects of slow and rapid
hemodialysis. Kidney Int 4: 177-187, 1973
16. Goldstein DA, Feinstein EI, Chui LA. et al: The rela-
tionship between the abnormalities in EEG and blood levels of
parathyroid hormone in dialysis patients. J Clin Endocrinol
Metabol 51: 130- 134. 1980
17. Kennedy AC. Luke RG, Linton AL: Dialysis disequilib-
rium syndrome. in Shaldon S. Cook GC (eds): Renal Failure.
Oxford, UK, Blackwell Scientific Publications, 1964, pp
66-79
18. Rodrigo F. Shideman J, McHugh R, et al: Osmolality
changes during hemodialysis. Ann Intern Med 86:554- 561,
1977
19. Alfrey AC. Mishell JM. Burks J. et al: Syndrome of
dyspraxia and multifocal seizures associated with chronic
hemodialysis. Trans Am Soc ArtifIntern Organs 18:257-261.
1972
20. Flendrig JA. Kruis H. Das HA: Aluminum and dialysis
dementia. Lancet 1: 1235. 1976
21. Ward MK. Feest TG, Ellis HA, et al: Osteomalacic dial-
ysis osteodystrophy: Evidence for a water-borne aetiological
agent. probably aluminum. Lancet 1:841:845. 1978
22. Wing AJ: Dialysis dementia in Europe: Report from the
registration committee of the EDTA. Lancet 2: 190-192, 1980
23. Pierides AM, Edwards WG. Cullum UX, et al:
Hemodialysis encephalopathy with osteomalacic fractures and
muscle weakness. Kidney Int 18: 115-124, 1980
24. Prior JC, Cameron EC, Knickerbocker WJ, et al:
MAHONEY AND ARIEFF
Dialysis encephalopathy and osteomalacic bone disease. Am J
Med 72:33-42, 1982
25. Dunea G, Mahurkar SD. Mamdami B. et al: Role of
aluminum in dialysis dementia. Ann Intern Med 88:502-504.
1978
26. Parkinson IS. Ward MK . Feest TG, et al: Fracturing
dialysis osteodystrophy and dialysis encephalopathy. Lancet
1:406-409, 1979
27. Rotundo A. Nevins TE. Lipton M, et al: Progressive
encephalopathy in children with chronic renal insufficiency in
infancy. Kidney Int 21:486-491, 1982
28. Alfrey AC. LeGendre GR. Kaehney WD: The dialysis
encephalopathy syndrome: Possible aluminum intoxication. N
Engl J Med 294: 184~ 188, 1976
29. Dewberry FL. McKinney TD, Stone WJ: The dialysis
dementia syndrome: Report of fourteen cases and review of the
literature. ASAIO 3: 102~ 108. 1980
30. Berkseth RO, Shapiro FL: An epidemic of dialysis en-
cephalopathy and exposure to high aluminum dialysate, in
Schreiner GE. Winchester JF (eds): Controversies in Nephrol-
ogy (vol II). Washington, DC. Georgetown University Press
1980, pp 42-51
31. Kaehney WD, Alfrey AC. Holman RE, et al: Aluminum
transfer during hemodialysis. Kidney Int 12:361-365. 1977
32. McDermott JR, Smith AI. Ward MK. et al: Brain alumi-
num concentration in dialysis encephalopathy. Lancet 1:901-
904. 1978
33. Weiss B: The behavioral toxicology of metals. Federa-
tion Proc 37:22-27,1978
34. Cartier F. Allain P, Gary J. et al: Progressive myoclonic
encephalopathy in dialysis patients. Nouv Presse Med 7:97-
102. 1978
35. McKinney TD, Basinger M. Dawson E, et al: Serum
aluminum levels in dialysis dementia. Nephron (in press)
36. Mayor GH, Sprague SM, Hourani MR, et al:
Parathyroid hormone mediated aluminum deposition and egress
in the rat. Kidney Int 17:40-44. 1980