Download Posterior Pituitary Dysfunction after Traumatic Brain Injury

0021-972X/04/$15.00/0
Printed in U.S.A.
The Journal of Clinical Endocrinology & Metabolism 89(12):5987–5992
Copyright © 2004 by The Endocrine Society
doi: 10.1210/jc.2004-1058
Posterior Pituitary Dysfunction after Traumatic
Brain Injury
AMAR AGHA, EVAN THORNTON, PATRICK O’KELLY, WILLIAM TORMEY, JACK PHILLIPS,
CHRISTOPHER J. THOMPSON
AND
Academic Departments of Endocrinology (A.A., E.T., C.J.T.), Renal Medicine (P.O.), Clinical Chemistry (W.T.), and
Neurosurgery (J.P.), Beaumont Hospital, Dublin 9, Ireland
Disorders of water balance are well recognized after traumatic brain injury (TBI), but there are no reliable data on
their true prevalence in post-TBI patients. We aimed to evaluate the prevalence of posterior pituitary dysfunction in a
large cohort of survivors of TBI.
One hundred two consecutive patients (85 males) who suffered severe or moderate TBI were evaluated for diabetes insipidus (DI) at a median of 17 months (range 6 –36 months) after
the event, using the 8-h water deprivation test (WDT). Their
results were compared against normative data obtained from 27
matched, healthy controls. Patients’ medical records were retrospectively reviewed for the presence of abnormalities of salt
and water balance in the immediate post-TBI period.
Twenty-two patients (21.6%) developed DI in the immediate
post-TBI period (acute DI group), of whom five had abnormal
WDT on later testing. In total, seven patients (6.9%) had abnor-
T
RAUMATIC BRAIN INJURY (TBI) is the leading cause
of death and disability in young adults (1, 2). Disorders
of salt and water balance are the most commonly recognized
medical complications in the immediate post-TBI period (3)
and contribute to the early morbidity observed in TBI patients. Most cases of acute posttraumatic diabetes insipidus
(DI) are transient, and traditionally permanent DI after head
trauma has been regarded as rare (4). However, patients with
partial deficiency of arginine vasopressin (AVP) can be easily
missed because they may have less severe symptoms and
their posttraumatic clinical course is often complicated by
significant neurological and cognitive disabilities.
Posterior pituitary function in survivors of TBI remains
poorly investigated. There are no reliable data on the prevalence of permanent posttraumatic DI. In this study, we
assessed posterior pituitary function in a large cohort of
patients after moderate or severe TBI.
Patients and Methods
Patients
One hundred two TBI patients (85 males), aged 15– 65 yr (median, 28
yr; mean ⫾ sd, 32.9 ⫾ 14.7 yr) who were admitted to the neurosurgical
Abbreviations: AVP, Arginine vasopressin; BMI, body mass index;
CI, confidence interval; CT, computerized tomography; DBI, diffuse
brain injury; DI, diabetes insipidus; GCS, Glasgow Coma Scale; ICP,
intracranial pressure; ITU, intensive care unit; SIADH, syndrome of
inappropriate secretion of antidiuretic hormone; TBI, traumatic brain
injury; WDT, water deprivation test.
JCEM is published monthly by The Endocrine Society (http://www.
endo-society.org), the foremost professional society serving the endocrine community.
mal WDT (permanent DI group), five of whom had partial DI.
Patients in the acute and permanent DI groups were more
likely to have more severe TBI, compared with the rest of the
cohort (P < 0.05). In the immediate post-TBI period, 13 patients (12.9%) had syndrome of inappropriate secretion of antidiuretic hormone, which persisted in one patient, and one
other patient developed cerebral salt wasting.
Diabetes insipidus and syndrome of inappropriate secretion of antidiuretic hormone were common in the immediate
post-TBI period. Permanent DI was present in 6.9% of patients
who survived severe or moderate TBI, which is higher than
traditionally thought. Identification of patients with partial
posttraumatic DI is important because appropriate treatment
may reduce morbidity and optimize the potential for
recovery. (J Clin Endocrinol Metab 89: 5987–5992, 2004)
unit in Beaumont Hospital between September 2000 and September 2002
and 27 matched healthy controls were included in the study (Table 1).
Beaumont Hospital is the national neurosurgical center for the Republic
of Ireland and has a catchment area of 3.5 million people. Patients were
identified from the Beaumont Hospital head trauma database. The database contains demographic and mortality data and information about
the nature and severity of the TBI but no information about any medical
complications or drug therapy. Patients were eligible for inclusion in the
study if they suffered severe or moderate TBI (see below), were between
15 and 65 yr of age, and were at least 6 months post the acute injury.
Exclusion criteria were: pregnant women, patients with established renal disease, patients with raised creatinine (⬎120 ␮mol/liter), patients
on lithium or other medications known to cause renal insensitivity to
AVP, patients with diabetes mellitus with hemoglobin A1C greater than
6.5%, and patients with hypokalemia or hypercalcemia. One hundred
twenty-eight patients were eligible for inclusion in the study. Twenty-six
patients were excluded for the following reasons: four died since leaving
the hospital, 13 left Ireland or were uncontactable, four were too ill to
participate, and five declined to participate. Two of the participating
patients, who were blindly selected from the database, were taking oral
desmopressin treatment. They were discharged from the hospital on
desmopressin following their TBI but were not previously investigated
by means of the water deprivation test (WDT).
All patients had suffered severe or moderate head trauma according
to the initial postresuscitation and presedation Glasgow Coma Scale
(GCS) score (Table 1) (5). Fifty-seven patients (55.9%) had documented
severe head injury, as defined by a GCS score of 8/15 or less, and 42
(41.2%) had documented moderate injury, defined by a GCS score of
9/15 to13/15 (6). The three remaining patients had no documented GCS
score but were admitted to the intensive care unit (ITU) and therefore
were assumed to have a GCS score of 13 or less. The median GCS score
was 8/15.
The causation of TBI was road traffic accidents in 44 patients, falls in
30 patients, alleged assault in 13 patients, and other mechanisms in 15
patients. Assessment of outcome post TBI was done using the Glasgow
Outcome Scale score (7). All patients, except one, had computerized
tomography (CT) evidence of brain injury (Table 1). CT appearance was
5987
Downloaded from jcem.endojournals.org on March 21, 2006
5988
J Clin Endocrinol Metab, December 2004, 89(12):5987–5992
Agha et al. • Diabetes Insipidus and Head Injury
TABLE 1. Baseline characteristics of patients and controls
Age (yr)
Male:female ratio
BMI (kg/m2)
GCS score at time TBI, n (%)
ⱕ8/15
9/15–13/15
Unknown
CT scan appearance, n (%)
FBI
DBI
Normal
Cerebral edema
Patients
(n ⫽ 102)
Controls
(n ⫽ 27)
P
32.9 ⫾ 14.8
85:17
24.5 ⫾ 4.1
33.6 ⫾ 9.6
21:6
24.7 ⫾ 4.4
0.365
0.573
0.805
57 (56)
42 (41)
3 (3)
86 (84)
15 (15)
1 (1)
72 (71)
Results are expressed as means ⫾
SD.
FBI, Focal brain injury.
classified as showing focal brain injury (extradural, subdural, or intracerebral hematomas) or diffuse brain injury (DBI; shearing axonal injury), with or without CT evidence of raised intracranial pressure (ICP)
(8). Eighty-six patients (84.3%) had focal brain injury and 15 patients
(14.7%) had DBI. Seventy-two patients (70.6%) had CT evidence of raised
ICP. Fifty-five patients (54%) underwent operative mass evacuation.
Eighty-four patients (82.4%) had intracranial pressure monitoring and
were sedated and ventilated in the neurosurgical ITU. The duration of
ITU stay was 14 ⫾ 10 d. Thirty-two patients (31.4%) had additional
non-TBI trauma.
Methods
Patients were tested at a median of 17 months after injury (range 6 –36
months). Testing was performed in the pituitary investigation day ward.
Testing started at 0800 h. Undiluted blood samples were drawn from a
heparinized cannula inserted in an antecubital vein.
The WDT (9) was carried out in all patients and controls. Fluids were
allowed ad libitum until 0700 h. All subjects had been advised to avoid
alcohol for 48 h and nicotine and caffeine for 12 h before the study.
Patients were asked to void their bladders at 0700 h; no fluid intake was
allowed after this until 1600 h. The two patients on desmopressin were
asked to withhold treatment on the evening before, and the morning of,
the test.
Testing started at 0800 h. Plasma and urine osmolalities, urine volume, thirst score, blood pressure, and weight were measured at time
0800, 1000, 1200, 1400, and 1600 h. Plasma sodium was measured at 0800
and 1600 h. All studies were medically supervised, with the intention
that studies would be stopped if subjects lost more than 5% of their body
weight. At the conclusion of the test, patients were allowed to drink
freely, and the volume of water drunk in 30 min was noted. Baseline
serum samples were also obtained for measurement of urea, creatinine,
calcium, and potassium.
Patients’ medical records were retrospectively reviewed for evidence
of salt and water abnormalities in the acute post-TBI phase, defined as
the time period from admission to the hospital until discharge to either
a rehabilitation institute or home. All 102 patients had their anterior
pituitary function assessed separately, and these results have been reported separately (10).
Definitions of abnormalities
WDT. The normal homeostatic response to dehydration is an increase in
plasma osmolality (plasma sodium) and thirst appreciation. The increase in plasma osmolality, or more precisely in plasma sodium, triggers vasopressin release, which acts on the distal renal tubule to stimulate water resorption, resulting in reduced urine output and increased
urine osmolality. However, in the presence of vasopressin deficiency or
renal resistance to vasopressin action (i.e. DI), the renal concentrating
ability is impaired and hypotonic polyuria results despite dehydration.
This is the underlying rationale of using the WDT to diagnose DI.
In practice, however, there is no consensus regarding the exact definition of a normal response to water deprivation. Some authors define
a normal response as a peak 8-h urine osmolality greater than 700
mOsm/kg (11), others use the 800 mOsm/kg cut-off (9), and others use
a definition of peak urine to peak plasma osmolalities ratio of greater
than 2 as normal (12).
In view of the lack of consensus in the literature, we studied 27
healthy controls during WDT. Twenty-four healthy controls (89%) had
peak urine osmolality exceeding 700 mOsm/kg. The three remaining
healthy controls had urine osmolalities of 642, 682, and 694 mOsm/kg;
however, all three had peak urine-plasma osmolalities ratio greater than
2 (2.2, 2.37, 2.32, respectively). Therefore, a normal WDT in our laboratory was defined as a peak urine osmolality greater than 700
mOsm/kg or a peak urine to plasma osmolalities ratio greater than 2.
In the immediate post-TBI period, DI was diagnosed if plasma sodium exceeded 145 mmol/liter in the presence of polyuria of more than
3.5 liter per 24 h and dilute urine (osmolality ⬍ 300 mOsm/kg) (13).
Syndrome of inappropriate secretion of antidiuretic hormone (SIADH)
was defined as: plasma osmolality less than 270 mOsm/kg, with a
corresponding urine osmolality of greater than 100 mOsm/kg and a spot
urine sodium more than 40 mmol/liter, in a euvolemic patient with
normal glucocorticoid secretion and thyroid function (14). Cerebral salt
wasting was defined as hypovolemic hyponatremia with diuresis and
natriuresis, not explained by renal dysfunction or diuretics (15).
Analytical methods
Plasma and urine osmolalities were measured by depression of the
freezing point method (2400 osmometer, Fiske, Norwood, MA). Plasma
sodium was measured using the ion-selective electrode method (Olympus 2700, Tokyo, Japan). Thirst was measured using a visual analog scale
previously shown to be accurate (16) and reproducible (17). Serum urea
and creatinine were measured by standard laboratory methods (Olympus 2700).
Statistical analysis
Prevalence figures are presented as percentages [95% confidence
intervals (CIs)]. Continuous data [age, body mass index (BMI), water
intake] were assessed for normality based on skewedness and kurtosis
and were also assessed for equality of variances between groups. As a
result of these tests, the data were log transformed before testing for
significance using the t test. However, for descriptive purposes, data are
expressed as untransformed mean ⫾ sd. GCS scores were analyzed
using a Wilcoxon rank-sum test for nonparametric measurements, and
categorical data were compared using the Fischer exact test. Multifactorial logistic regression models were developed to assess the effect of
appropriate variables in the presence of other confounding variables on
the development of DI and SIADH. The dependent variables for the
models were acute DI, permanent DI, and acute SIADH; the independent variables were age, gender, BMI, GCS score, cerebral edema, operative mass evacuation, and basal skull fractures.
Repeated measures ANOVA models were used to compare plasma
and urine osmolalities, plasma sodium, urine volume, and thirst at
different time intervals between groups. Multiple comparison tests,
using a Bonferroni correction factor, were used to determine whether
results reached significance.
The results were deemed significant for P ⬍ 0.05. Stata (version 8,
Stata Corp., College Station, TX) was used for the statistical analysis.
Ethics
The study was approved by the ethics section of Beaumont Hospital
Medical Research Committee. The purpose of the study was explained
carefully to patients and relatives, who were provided with written
information on the background to the study. After an interval of 1 wk,
patients who agreed to participate signed written consent for inclusion
in the study; written consent was given by next of kin, where
appropriate.
Results
Posterior pituitary dysfunction in the immediate
post-TBI period
In the acute post-TBI period, 22 patients (21.6%; 95% CI,
14.0 –30.8%) had evidence of DI (acute DI group) (Table 2).
Downloaded from jcem.endojournals.org on March 21, 2006
Agha et al. • Diabetes Insipidus and Head Injury
J Clin Endocrinol Metab, December 2004, 89(12):5987–5992 5989
Univariate comparisons between the acute DI and the rest of
patients are presented in Table 2. Multifactorial logistic regression analysis showed that acute DI was associated with
a lower GCS score (P ⫽ 0.043) and the presence of cerebral
edema (P ⫽ 0.045) but not related (P ⬎ 0.05) to age, gender,
BMI, operative mass evacuations, DBI, or basal skull fractures. There was a negative correlation between the GCS
score and the peak recorded plasma sodium (r ⫽ ⫺0.61, P ⫽
0.01).
Thirteen patients (12.7%; 95% CI, 7.0%–20.8%) had evidence of SIADH. None of the 13 patients was receiving
medications known to cause the syndrome. The median
plasma sodium was 128 mmol/liter (range 118 –131 mmol/
liter), and median blood urea was 4.3 mmol/liter (1.8 – 6.2
mmol/liter). The onset of SIADH was on d 1–2 in seven
patients (54%), d 4 –7 in five patients (38%), and one patient
(8%) developed SIADH on d 18 post TBI. All 13 patients had
normal glucocorticoid reserves and normal thyroid function.
Multifactorial logistic regression analysis showed that
SIADH was unrelated to age, BMI, gender, GCS score, the
presence of cerebral edema, operative mass evacuation, or
basal skull fracture (P ⬎ 0.05).
Hyponatremia due to cerebral salt wasting was diagnosed
in one patient only, a 28-yr-old man who developed hyponatremia of 120 mmol/liter on d 8 post TBI associated with
polyuria of 5.4 liters per 24 h, a natriuresis of 250 mmol/liter,
raised urea, and low central venous pressure.
Posterior pituitary dysfunction in the chronic phase of TBI
Seven patients (6.9%; 95% CI, 2.8 –13.6%) had abnormal
WDT (permanent DI group, Table 3), including both patients
treated for DI with oral desmopressin. All seven patients had
symptoms of polyuria, nocturia, and polydipsia. Five of
TABLE 2. Comparisons between patients with acute DI and rest
of cohort
a
Age (yr)
M:F ratio
BMI (kg/m2)a
GCS scoreb
Cerebral edema, n (%)
Mass evacuation, n (%)
Basal skull fracture, n (%)
Acute DI
(n ⫽ 22)
Acute non-DI
(n ⫽ 80)
P
28.8 ⫾ 11.9
17:5
24.2 ⫾ 5.4
6 (5–7)
20 (91)
11 (50)
4 (18)
34.1 ⫾ 15.3
68:12
24.5 ⫾ 3.7
8 (6 –12)
52 (65)
44 (55)
12 (15)
0.153
0.518
0.552
0.001
0.018
0.810
0.744
M, Male; F, female.
a
Results are expressed as mean ⫾ SD.
b
Results are expressed as median (interquartile range).
these had unequivocal evidence of DI in the immediate postTBI period; a sixth patient developed polyuria greater than
3 liters per 24 h on d 7 post TBI, although his plasma sodium
remained normal, and the seventh patient had normal
plasma sodium and urine output less than 2.5 per 24 h until
discharge from hospital on d 17 post TBI. The other 17 patients with acute posttraumatic DI had normal WDT (transient DI group). The remaining 78 patients had normal WDT
and no evidence of acute DI (non-DI group).
At the initiation of the WDT, patients with permanent DI
had a significantly higher plasma osmolality (median 298,
range 295–301 mOsm/kg) than healthy controls (median 287,
range 284 –291 mOsm/kg, P ⬍ 0.001). Figures 1 and 2 illustrate the changes in the measured variables during 8 h of
water deprivation in the four groups. Repeated-measures
ANOVA models showed that during 8 h water deprivation,
patients with permanent DI had higher plasma and lower
urine osmolalities, higher plasma sodium, higher urine volume, and higher thirst scores than patients in the transient
DI, non-DI, or control groups (P ⬍ 0.001 for each variable,
Figs. 1 and 2). All seven patients in the permanent DI group
had normal thirst appreciation, and they consumed more
water at the end of testing (1228 ⫾ 276 ml) than patients in
the non-DI group (563 ⫾ 152 ml, P ⬍ 0.001), the transient DI
group (581 ⫾ 128 ml, P ⬍ 0.001), or healthy controls (590 ⫾
125 ml, P ⬍ 0.001). All measured variables were not significantly different between patients in the transient DI, non-DI,
or control groups (Figs. 1 and 2).
Univariate comparisons between the permanent DI and
non-DI groups are presented in Table 4. Multifactorial logistic regression analysis showed that permanent DI was
associated with a lower GCS score (P ⫽ 0.043) but was not
statistically related (P ⬎ 0.05) to age, BMI, gender, DBI, the
presence of cerebral edema, operative mass evacuation, or
basal skull fractures. There was no relationship between the
presence of permanent DI and anterior pituitary dysfunction
(data not shown).
Two patients had evidence of SIADH. One had persistent
hyponatremia from the time of TBI, secondary to posttraumatic obstructive hydrocephalus, and the second patient had
developed SIADH after the acute TBI, when the selective
serotonin reuptake inhibitor citalopram was prescribed.
Discussion
We report the results of the largest study to date to examine the prevalence of posterior pituitary dysfunction in
survivors of TBI. There have been very few reliable data in
TABLE 3. Characteristics of the seven patients with permanent DI from a cohort of 102 with TBI
Subject/
gender
Age
(yr)
1/M
2/M
3/M
4/M
5/F
6/M
7/M
19
25
52
25
32
21
24
Cause of
TBI
RTA
RTA
RTA
RTA
Assault
RTA
Fall
GCS
score
Basal skull
fracture
Acute
DI
Peak Posm
(mOsm/kg)
Peak Uosm
(mOsm/kg)
5
13
7
3
5
6
4
N
N
Y
N
N
N
N
Y
Y
N
Y
Y
Y
N
311
305
305
314
308
307
304
165
520
467
80
480
506
521
Anterior pituitary
status
Normal
GTD
Normal
Normal
GHD, ACTHD
Normal
Normal
Posm, Plasma osmolality; Uosm, urine osmolality; RTA, road traffic accident; Y, yes; N, no; GHD, GH deficiency; GTD, gonadotropin
deficiency; ACTHD, ACTH deficiency; M, male; F, female.
Downloaded from jcem.endojournals.org on March 21, 2006
5990
J Clin Endocrinol Metab, December 2004, 89(12):5987–5992
Agha et al. • Diabetes Insipidus and Head Injury
FIG. 2. Changes in urine osmolality and two hourly urine volumes
during the WDT. Values are presented as means and SEM bars.
TABLE 4. Comparison between patients with permanent DI and
rest of cohort
a
Age (yr)
M:F ratio
BMI (kg/m2)a
GCS scoreb
Cerebral edema, n (%)
Mass evacuation, n (%)
Basal skull fracture, n (%)
Permanent DI
(n ⫽ 7)
Non-DI
(n ⫽ 95)
P
28.3 ⫾ 11.2
6:1
24.4 ⫾ 4.9
5 (4 –7)
5 (71)
4 (57)
1 (14)
33.3 ⫾ 15.0
79:16
24.5 ⫾ 4.1
8 (6 –11.5)
67 (71)
51 (53)
15 (16)
0.496
1.000
0.911
0.038
1.000
1.000
1.000
M, Male; F, female.
a
Results expressed as mean ⫾ SD.
b
Results expressed as median (interquartile range).
FIG. 1. Changes in plasma osmolality, thirst, and plasma sodium
during the WDT. Values are presented as means and SEM bars.
the literature concerning the prevalence of posttraumatic DI.
Our data show that frequency of permanent posttraumatic
DI is higher than previously thought (4). This can be explained by our finding that five of the seven DI patients had
partial defects, with mild symptoms, which had been clinically overlooked and gone undiagnosed.
Bohnen et al. (18) studied 38 patients, 5 wk after mild TBI.
They reported that eight (21%) had evidence of mild DI, a
much higher figure than in our cohort. However, the authors
measured early-morning plasma and urine osmolalities after
an unobserved overnight fast, a method that has been shown
to be inaccurate for the diagnosis of DI (19). Furthermore,
they defined DI on the basis of a plasma osmolality greater
than 295 and urine osmolality less than 1000 mOsm/kg,
Downloaded from jcem.endojournals.org on March 21, 2006
Agha et al. • Diabetes Insipidus and Head Injury
cut-offs that are too liberal and likely to result in overdiagnosis. In their review of published case report of posttraumatic hypopituitarism, Edwards and Clark (20) found DI to
occur in 23 of 53 case reports, with transient symptoms in
nine patients, although cases were selected for inclusion on
the basis that patients had hypopituitarism, introducing a
potential for bias. Our prevalence of DI is higher than that
reported in a retrospective study, Wong et al. (21), who found
DI in 3.7% of neurosurgical intensive care patients, although
they did not report on the frequency of DI in the subgroup
of TBI patients. We feel that our data, derived from a large,
unselected cohort studied with standard methodologies, are
more likely to reflect the true prevalence of persistent DI after
TBI.
The prevalence of acute posttraumatic DI in this study was
similar to that reported in a recent smaller prospective study
of 50 patients studied at the time of acute TBI (22). However,
in contrast to the last study, we found a significant association between acute posttraumatic DI and more severe TBI,
as indicated by lower GCS scores or CT evidence of raised
ICP due to cerebral edema. Naturally, patients with cerebral
edema tend to have lower GCS scores; therefore, the inherent
association between the two variables makes it more difficult
to statistically separate their independent effects on the development of acute DI. However, because both variables are
markers of the severity of TBI, our conclusion that acute DI
is associated with more severe trauma remains valid. This
conclusion is further supported by the finding of a negative
correlation between the GCS score and the peak recorded
plasma sodium, which is a surrogate marker for the severity
of DI. The findings in this study concur with anecdotal reports and clinical experience and may reflect the larger number of patients in our current study, which allowed more
valid statistical comparisons.
Five of the 22 acute DI patients were found to have persistent (permanent) DI, whereas the remaining 17 had no
evidence of any abnormalities in any of the variables measured, indicating complete recovery of posterior pituitary
function. Interestingly, two of the seven patients with permanent DI had no definite acute abnormalities. It is possible
that their partial DI was masked by adequate hydration with
iv fluids or drinking in the acute post-TBI period. Alternatively, DI may have developed in the postacute phase, which
has been previously reported (23). Two of the seven patients
with permanent DI had been on desmopressin therapy before the commencement of the study. However, the investigators were unaware of this fact, and the two patients were
not preselected. Both patients had severe DI with good clinical response to desmopressin. The other five patients had
milder symptoms and were considered to have partial AVP
deficiency (12), which has previously been reported after TBI
(24). Although the criteria that we used to define DI using the
WDT may also be met by some subjects with primary polydipsia, the finding of high plasma osmolalities, both at the
initiation and the conclusion of the WDT in all patients with
abnormal WDT, confirms the diagnosis of DI.
Patients who developed permanent DI were more likely to
have lower GCS scores, indicating more severe injury. In
contrast to data from the immediate post-TBI period, there
was no relationship between the presence of cerebral edema
J Clin Endocrinol Metab, December 2004, 89(12):5987–5992 5991
and the development of permanent DI. Such an association
cannot be excluded because the small number of subjects in
the permanent DI group may have precluded reliable statistical comparisons. Another possible explanation is that
transient DI may be related, in some cases, to edema around
the hypothalamic-pituitary region, which later resolves, with
resumption of normal physiological AVP secretion. In contrast, permanent DI may result from direct irreversible damage to the hypothalamic paraventricular and supraoptic neurones or their projection to the posterior pituitary.
Twenty-eight percent of this cohort had evidence of one or
more of anterior pituitary hormone deficiency; however, we
found no association between anterior hypopituitarism and
acute or chronic posterior pituitary dysfunction.
In this study, we did not measure plasma vasopressin
levels. Plasma vasopressin level can be useful in differentiating patients with cranial DI from those with nephrogenic
DI (25). However, we were careful to exclude any patients
with medical conditions or medications known to cause
nephrogenic DI. In addition, five of the seven patients with
permanent DI had clear evidence of DI presenting acutely in
the immediate post-TBI period, and all seven patients responded clinically to treatment with desmopressin, with concentration of urine to more than 700 mOsm/kg. We are
confident that all seven patients with DI had vasopressin
deficiency; and because the 95 patients classified as normal
had unequivocally normal WDT, with all measured variables
similar to healthy controls, it is highly unlikely that AVP
measurement would have altered the diagnosis in any individual studied.
Posttraumatic DI may result from inflammatory edema
around the hypothalamus or posterior pituitary, with resolution as the swelling resolves. It can also result from direct
damage to the paraventricular and supraoptic hypothalamic
neurones, the pituitary stalk, or axon terminals in the posterior pituitary. These abnormalities can be transient if the
supraoptic and paraventricular neurones form new vascular
connections or become permanent (4).
Posttraumatic SIADH is caused by uncontrolled release of
AVP as a result of damage to the pituitary stalk or the posterior pituitary (26). Acute hyponatremia secondary to
SIADH was present in 12.7% of patients. Studies on the
prevalence of SIADH post-TBI have yielded conflicting data,
with figures ranging from 2.3 to 36.6% (22, 27–31). The conflicting figures in the literature may reflect different selection
criteria for the cohorts of patients studied, different criteria
used to define SIADH, and varying duration of monitoring
after TBI. SIADH persisted in only one patient, who had
obstructive hydrocephalus. Glucocorticoid deficiency was
excluded by dynamic tests of the hypothalamo-pituitaryadrenal axis in all cases. We are therefore confident that none
of the patients with SIADH in this cohort had undiagnosed
glucocorticoid deficiency. Cerebral salt-wasting syndrome is
a rare complication of head trauma, which has been reported
previously (15), but we were able to diagnose with confidence in only one patient.
Untreated DI leads to polyuria. In the early post-TBI period, water intake may be inadequate to compensate for the
polyuria because of impaired cognition, physical disability,
or coexistent hypodipsia. This can lead to dehydration and
Downloaded from jcem.endojournals.org on March 21, 2006
5992
J Clin Endocrinol Metab, December 2004, 89(12):5987–5992
a hyperosmolar state with impairment of rehabilitation and
increased morbidity. This is particularly a problem with adipsic diabetes insipidus; thirst appreciation is almost uniformly normal in DI (32), but adipsic DI has been reported
in association with TBI (33) and is associated with poor
prognosis.
In conclusion, we have shown, in a large cohort of survivors of TBI, a significant frequency of early and late posterior
pituitary dysfunction. Recognition of these treatable abnormalities is important, to maximize the potential for recovery,
in this high morbidity group of patients. Therefore, all patients presenting with acute posttraumatic DI should be reevaluated in the postacute phase using the WDT. Patients
without a recognized history of acute posttraumatic TBI and
who have no symptoms of thirst, polyuria, or nocturia and
a 24-h urine output of less than 3 liters need no further
assessment. However, in the presence of symptoms or polyuria of more than 3 liters/d, a WDT should be performed.
Acknowledgments
The authors thank Professor Dermot Kenny and the staff of the Royal
College of Surgeons in Ireland Clinical Research Centre (Dublin, Ireland), at which part of this study was conducted. The authors are
indebted to Sister Siobhan Lydon and her staff for their assistance with
the study.
Agha et al. • Diabetes Insipidus and Head Injury
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Received June 4, 2004. Accepted September 20, 2004.
Address all correspondence and requests for reprints to: Dr. Christopher J. Thompson, Department of Endocrinology, Beaumont Hospital,
Beaumont Road, Dublin 9, Ireland. E-mail: [email protected].
This work was supported by an unrestricted educational grant, obtained in open competition, from Pfizer International Research Grants.
25.
26.
References
27.
1. Klasbeek WD, McLaurin, RL, Harris BSH, Miller JD 1980 The national head
and spinal cord survey findings. J Neurosurg 53:519 –531
2. Van Baalen B, Odding E, Maas AIR, Ribbers GM, Bergen MP, Stam HJ 2003
Traumatic brain injury: classification of initial severity and determination of
functional outcome. Disabil Rehabil 25:9 –18
3. Kaufman HH, Timberlake G, Voelker J, Pait TG 1993 Medical complications
of head injury. Med Clin North Am 77:43– 60
4. Yaun X-Q, Wade CE 1991 Neuroendocrine abnormalities in patients with
traumatic brain injury. Front Neuroendocrinol 12:209 –230
5. Teasdale G, Jennet B 1974 Assessment of coma and impaired consciousness:
a practical scale. Lancet 2:81– 84
6. World Health Organization 1980 International classification of impairments,
disabilities, and handicaps. Geneva: World Health Organization
7. Jennett B, Bond M 1975 Assessment of outcome after severe brain damage.
Lancet 1:480 – 485
8. Teasdale G, Teasdale E, Hadley D 1992 Computed tomographic and magnetic
28.
29.
30.
31.
32.
33.
resonance imaging classification of head injury. J Neurotrauma 9(Suppl 1):
249 –257
Baylis PH 1998 Diabetes insipidus. J R Coll Physicians Lond 32:108 –111
Agha A, Rogers B, Tormey W, Phillips J, Thompson CJ, Prevalence of hypopituitarism in survivors of traumatic brain injury. J Clin Endocrinol Metab,
in press
Moore K, Thompson CJ, Trainer P 2003 Disorders of water balance. Clin Med
3:28 –33
Dashe AM, Cramm RE, Crist CA, Habener J F, Solomon DH 1963 A water
deprivation test for the differential diagnosis of polyuria. JAMA 185:699 –703
Seckl J, Dunger D 1989 Post-operative diabetes insipidus. BMJ 298:2–3
Verbalis JG 1989 Hyponatraemia: review. Baillieres Clin Endocrinol Metab
3:499 –530
Smith DM, McKenna K, Thompson CJ 2000 Hyponatraemia. Clin Endocrinol
(Oxf) 52:667– 678
Thompson CJ, Bland J, Burd J, Baylis PH 1986 The osmotic thresholds for
thirst and vasopressin are similar in healthy man. Clin Sci (Lond) 71:651– 656
Thompson CJ, Selby P, Baylis PH 1991 Reproducibility of osmotic and nonsomotic tests of vasopressin secretion in man. Am J Physiol 260:R533–R539
Bohnen N, Twijnstra A, Jolles J 1993 Water metabolism and post-concussional
symptoms 5 weeks after mild head injury. Eur Neurol 33:77–79
Robertson GL 1985 Diagnosis of diabetes insipidus. Front Horm Res 13:176 –
189
Edwards OM, Clark JDA 1986 Post-traumatic hypopituitarism. Medicine (Baltimore) 65:281–290
Wong MF, Chin NM, Lew TW 1998 Diabetes insipidus in neurosurgical
patients. Ann Acad Med Singapore 27:340 –343
Agha A, Roger B, Mylotte D, Taleb F, Tormey W, Phillips J, Thompson CJ
2004 Neuroendocrine dysfunction in the acute phase of traumatic brain injury.
Clin Endocrinol (Oxf) 60:584 –591
Alaca R, Yilmaz B, Gunduz S 2002 Anterior hypopituitarism with unusual
delayed onset of diabetes insipidus after penetrating head injury. Am J Phys
Med Rehabil 81:788 –791
Notman DD, Mortek MA, Moses AM 1980 Permanent diabetes insipidus
following head trauma: observations on ten patients and an approach to
diagnosis. J Trauma 20:599 – 602
Baylis PH, Thompson CJ 2001 Diabetes insipidus and hyperosmolar syndromes. In: Becker KL, ed. Principles and practice of endocrinology and metabolism. 3rd ed. Philadelphia: Lippincott, Williams & Wilkins; 285–293
Verbalis JG 2001 Inappropriate antidiuresis and other hypoosmolar states. In:
Becker KL, ed. Principles and practice of endocrinology and metabolism. 3rd
ed. Philadelphia: Lippincott, Williams & Wilkins; 293–305
Becker LH, Daniel RK 1973 increased antidiuretic hormone production after
trauma due to craniofacial complex. J Trauma 13:112–115
Twijnstra A, Minderhoud JM 1980 Inappropriate secretion of antidiuretic
hormone in patients with head injuries. Clin Neurol Neurosurg 82:263–268
Doczi T, Tarjanyi J, Huszka E, Kiss J 1982 Syndrome of inappropriate secretion of antidiuretic hormone (SIADH) after head injury. Neurosurgery
10:685– 688
Born JD, Hans P, Smitz S, Legros JJ, Kay S 1985 Syndrome of inappropriate
secretion of antidiuretic hormone after severe head injury. Surg Neurol 23:
383–387
Vigerhoets F, de Tribolet N 1988 Hyponatremia hypo-osmolality in neurosurgical patients. “Appropriate secretion of ADH” and “cerebral salt wasting
syndrome.” Acta Neurochir (Wien) 91:50 –54
Thompson CJ, Baylis PH 1987 Thirst in diabetes insipidus: clinical relevance
of quantitative assessment. Q J Med 65:853– 862
Smith D, McKenna K, Moore K, Tormey W, Finucane J, Phillips J, Baylis P,
Thompson CJ 2002 Baroregulation of vasopressin release in adipsic diabetes
insipidus. J Clin Endocrinol Metab 87:4564 – 4568
JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.
Downloaded from jcem.endojournals.org on March 21, 2006