S P E C I A L F E A T U R E

R e v i e w

Endocrine Responses to Critical Illness: Novel Insights

and Therapeutic Implications

Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
Eva Boonen and Greet Van den Berghe
Clinical Division and Laboratory of Intensive Care Medicine, Department of Cellular and Molecular
Medicine, KU Leuven, B-3000 Leuven, Belgium

Context: Critical illness, an extreme form of severe physical stress, is characterized by important
endocrine and metabolic changes. Due to critical care medicine, survival from previously lethal
conditions has become possible, but many patients now enter a chronic phase of critical illness. The
role of the endocrine and metabolic responses to acute and prolonged critical illness in mediating
or hampering recovery remains highly debated.

Evidence Acquisition: The recent literature on changes within the hypothalamic-pituitary-thyroid

axis and the hypothalamic-pituitary-adrenal axis and on hyperglycemia in relation to recovery from
critical illness was critically appraised and interpreted against previous insights. Possible thera-
peutic implications of the novel insights were analyzed. Specific remaining questions were
formulated.

Evidence Synthesis: In recent years, important novel insights in the pathophysiology and the
consequences of some of these endocrine responses to acute and chronic critical illness were
generated. Acute endocrine adaptations are directed toward providing energy and substrates for
the vital fight-or-flight response in a context of exogenous substrate deprivation. Distinct endo-
crine and metabolic alterations characterize the chronic phase of critical illness, which seems to be
no longer solely beneficial and could hamper recovery and rehabilitation.

Conclusions: Important novel insights reshape the current view on endocrine and metabolic re-
sponses to critical illness and further clarify underlying pathways. Although many issues remain
unresolved, some therapeutic implications were already identified. More work is required to find
better treatments, and the optimal timing for such treatments, to further prevent protracted
critical illness, to enhance recovery thereof, and to optimize rehabilitation. (J Clin Endocrinol
Metab 99: 1569 –1582, 2014)

ritical illness is defined as any life-threatening condi-
C tion requiring support of vital organ functions to
prevent imminent death. This condition can be evoked by
a variety of insults such as multiple trauma, complicated
surgery, and severe medical illnesses. Without modern
critical care medicine, critically ill patients would not sur-
vive. Critical illness is thus the ultimate form of severe
physical stress, and all the immediate biological responses
that are evoked are expected to be of greater magnitude in
critically ill patients. These immediate stress responses
comprise many orchestrated endocrine adaptations that
are presumably directed toward providing the required
energy for the fight-or-flight response in a context of ex-
ogenous substrate deprivation. Indeed, alterations within
the different hypothalamic-pituitary axes bring about li-
polysis, proteolysis, and gluconeogenesis and redirect en-
ergy consumption toward those processes that mediate
acute survival, whereas anabolism is postponed to more
prosperous times.
Although survival from previously lethal conditions is
nowadays possible, often recovery does not swiftly follow,
and patients enter a chronic phase of critical illness during

ISSN Print 0021-972X ISSN Online 1945-7197
Printed in U.S.A.
Copyright © 2014 by the Endocrine Society
Received November 15, 2013. Accepted February 6, 2014.
First Published Online February 11, 2014
Abbreviations: CBG, cortisol binding globulin; D1, type 1 deiodinase; D2, type 2 deiodi-
nase; D3, type 3 deiodinase; GR, glucocorticoid receptor; ICU, intensive care unit; SIRS,
systemic inflammatory response syndrome; TR, thyroid hormone receptor.

doi: 10.1210/jc.2013-4115 J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582 jcem.endojournals.org 1569
1570 Boonen and Van den Berghe Novel Insights on Endocrine Changes in the ICU J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582

which they continue to depend upon vital organ support

for weeks, whereas the original trigger of the critical illness
has long been resolved. This stage is characterized by dis-
tinct endocrine and metabolic alterations that may no lon-
ger be solely beneficial because they may hamper recovery.
An example is the relative maintenance of fat stores while

Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
large amounts of proteins continue to be wasted from skel-
etal muscle and organs (1). This response may impair re-
covery of vital organ functions, extend weakness, hamper
rehabilitation (2), and expose patients to severe, often in-
fectious, complications (3). The understanding of the
mechanisms determining why certain patients recover and
others don’t remains very limited, but recent studies point
to variable abilities to remove cell damage as playing a key
role (4, 5). When patients remain dependent upon critical
care support, it is ultimately decided to withdraw care
because of futility. Hence, further understanding the un-
derlying pathways of recovery and investigating whether
these pathways can be beneficially affected by treatment is
of high clinical relevance.
In recent years, important novel insights in the patho-
physiology and the consequences of these endocrine re-
sponses to critical illness were generated. This review sum-
marizes these insights with a specific focus on the
hypothalamic-pituitary-thyroid axis, the hypothalamic-
pituitary-adrenal axis, and the impact of the hyperglyce-
mic response on recovery from critical illness. Any thera-
peutic implications of these novel insights are critically
analyzed. Figure 1. Changes in the central and peripheral thyroid axis in acute
vs prolonged critical illness. The top panel shows reduced TRH gene
expression in the hypothalamus of prolonged ill patients. The second
panel illustrates adaptations in nocturnal TSH secretion with a loss of
Hypothalamic-Pituitary-Thyroid Axis pulsatility during prolonged critical illness. The lower panels summarize
schematically the changes in circulating thyroid hormone
Responses within the hypothalamic-pituitary- concentrations and changes in peripheral deiodinase enzyme activity
thyroid axis during acute critical illness levels. [Figure was drafted from original data in E. Fliers et al:
Decreased hypothalamic thyrotropin-releasing hormone gene
It has long been known that both fasting and acute expression in patients with nonthyroidal illness. J Clin Endocrinol
illnesses immediately affect circulating levels of thyroid Metab. 1997;82:4032– 4036 (26). © The Endocrine Society. Y.
hormones. Most typically, plasma concentrations of T3 Debaveye et al: Regulation of tissue iodothyronine deiodinase activity
in a model of prolonged critical illness. Thyroid. 2008;18:551–560 (46),
decrease and plasma concentrations of rT3 rise, suggesting with permission. © American Thyroid Association. I. Vanhorebeek et al:
an immediate inactivation of thyroid hormone in periph- Endocrine aspects of acute and prolonged critical illness. Nat Clin Pract
eral tissues such as the liver, likely mediated by a sup- Endocrinol Metab. 2006;2:20 –31 (125), with permission. © Macmillan
Publishers Limited. L. Mebis et al: Thyroid axis function and dysfunction
pressed activity of the type 1 deiodinase (D1) and/or an in critical illness. Best Pract Res Clin Endocrinol Metab. 2011;25:745–
activated type 3 deiodinase (D3) (6, 7). Concentrations of 757 (126), with permission. © Elsevier Ltd.
T4 and TSH have been shown to be briefly increased im-
mediately after surgery (7). Thereafter, plasma TSH and Several possible mediators of the acute fall in plasma T3
T4 concentrations often return to “normal,” although a concentrations in critically ill patients include the lack of
normal nocturnal TSH surge is absent (8, 9). This con- nutrients, the release of cytokines, or hypoxia (10 –12).
stellation of low plasma T3 concentrations and elevated TNF-␣, IL-1, and IL-6 are capable of mimicking the acute
rT3 is generally referred to as the acute low-T3 syndrome, stress-induced alterations within the thyroid axis. How-
the euthyroid-sick syndrome, or the nonthyroidal illness ever, neutralizing antibodies to these cytokines in a human
syndrome (Figure 1). experiment of lipopolysaccharide-induced inflammation
doi: 10.1210/jc.2013-4115
failed to restore normal thyroid hormone concentrations
(13). Acute decreases in plasma concentrations of thyroid
hormone binding proteins and the inhibition of hormone
binding, transport, and metabolism by elevated levels of
free fatty acids and bilirubin may also play a role (14).
The low T3 concentrations that occur with fasting have
been shown to be adaptive because they appear to protect
the organism against the deleterious catabolic conse-
quences of a lack of macronutrients (15, 16). In critical
illness, it was suggested that the low T3 concentrations
could be maladaptive because the magnitude of the acute
T3 decrease was associated with the severity of illness and
with the risk of death (17, 18). However, the acute fall in
circulating levels of thyroid hormone in response to illness
could also be an adaptive attempt to reduce energy ex-
penditure, as happens with fasting in healthy subjects, in
which case it should be left untreated (15). Improved post-
operative cardiac function was observed after short-term
iv administration of T3 to patients during elective cardiac
surgery (19, 20). However, supranormal plasma T3 con-
centrations were evoked, and thus it is uncertain whether
these findings were merely due to a pharmacological ef-
fect. Recently, the results of a large randomized controlled
trial investigating the impact of early parenteral nutrition,
as compared with tolerating pronounced caloric deficit in
critically ill patients, provided indirect evidence for an
adaptive nature of the low T3 levels (21, 22). This study
revealed that providing nutrition in the acute phase of
critical illness impaired rather than improved outcome.
The provision of macronutrients partially prevented the
acute thyroid hormone changes, which were also recently
observed in a rabbit model of critical illness (23). In the
clinical study, specifically, the rise in T3 and in the ratio of
T3 over rT3 with early forceful feeding statistically ex-
plained the worsening of the outcome (22). These data
therefore suggest that at least part of the acute fall in T3
concentrations during critical illness is related to the con-
comitant fasting, and that this part of the response is likely
adaptive. Benefits include the expected reduction in energy
expenditure with low T3 levels, or a direct effect of in-
creased D3 activity locally in granulocytes, which could
optimize bacterial killing capacity (12, 24).
Responses within the hypothalamic-pituitary-
thyroid axis during prolonged critical illness
However, when patients are treated in intensive care
units (ICUs) for several weeks, receiving full enteral and/or
parenteral nutrition, the alterations within the thyroid
axis appear different. In this phase of critical illness, low
plasma T3 concentrations now coincide with low T4 con-
centrations and low-normal TSH concentrations in a sin-
gle morning sample (25). Moreover, overnight repeated
jcem.endojournals.org 1571
sampling revealed that the pulsatility of TSH secretion is
virtually lost, which relates to low plasma thyroid hor-
mone levels, a presentation resembling central hypothy-
roidism (Figure 1) (25). In line with this interpretation,
Fliers et al (26) demonstrated in postmortem brain sam-
ples of chronic critically ill patients that the gene expres-
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
sion of TRH in the hypothalamic paraventricular nuclei
was much lower than in patients who died after acute
insults (Figure 1). Furthermore, a positive correlation was
observed between the TRH mRNA expression and the
plasma concentrations of TSH and T3. Together, these
data indicate that production and/or release of thyroid
hormones is reduced in prolonged critical illness due to
reduced hypothalamic stimulation of the thyrotropes, in
turn leading to reduced stimulation of the thyroid gland.
The observation that a rise in TSH levels precedes the onset
of recovery from severe illness further supports this inter-
pretation (27).
The factors triggering hypothalamic suppression dur-
ing prolonged critical illness are unknown. Because
plasma cytokine concentrations are usually much lower in
the prolonged phase of critical illness (28), other mecha-
nisms likely play a role, like endogenous dopamine or el-
evated cortisol levels in the hypothalamus, because exog-
enous dopamine and hydrocortisone are known to
provoke or aggravate hypothyroidism in critical illness
(29 –31). A local increase in type 2 deiodinase (D2) activity
in the hypothalamus could elevate local thyroid hormone
levels, whereby the setpoint for feedback inhibition could
be altered (32). Indeed, in a rabbit model of prolonged
critical illness and low thyroid hormone plasma concen-
trations, hypothalamic TRH mRNA was low and D2
mRNA was high. However, the hypothalamic T4 and T3
concentrations were not increased (33). Increased pitu-
itary D2 could also play a role in suppressing local TSH
mRNA (34), although this was not confirmed in an animal
model of prolonged critical illness (35).
During prolonged critical illness, peripheral tissues
seem to respond to low T3 levels to increase local hormone
availability and effects. For example, in skeletal muscle
and liver biopsies from prolonged critically ill patients, the
monocarboxylate transporter MCT-8 was overexpressed
(Figure 2) (36). This was confirmed in an animal model,
where the up-regulation of the monocarboxylate trans-
porters in liver and kidney was reversible by treatment
with thyroid hormones (36, 37). Also, in skeletal muscle
biopsies from prolonged critically ill patients, D2 expres-
sion and activity were up-regulated as compared with
healthy controls and with acutely ill patients (Figure 1)
(37). Up-regulation of D2 in lungs was recently found to
be adaptive in sepsis and acute lung injury, further accen-
tuated by the observation that a D2 polymorphism was
1572 Boonen and Van den Berghe Novel Insights on Endocrine Changes in the ICU J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582

longed critically ill patients (39). Together, the data sug-

Plasma Hormone Concentrations gest that when the production of thyroid hormones falls in
160
2.0 prolonged critical illness, peripheral tissues adapt by in-
T3 (nmol/l)

120 creasing thyroid hormone transporters, local activation of

T4 (nmol/l)
1.5
thyroid hormone, and gene expression of the active recep-
1.0 * 80 * tor isoform.
HUMAN PATIENTS

Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
0.5 40 In protracted critical illness, low T3 levels were found to
0.0 0 correlate inversely with markers of muscle breakdown and
of bone loss, which could indicate either an adaptive and
Tissue Expression protective response against catabolism or a causal mal-
adaptive relationship (40). Because the cause of the low
Liver Muscle
2.0 * 12 thyroid hormone levels during prolonged critical illness
*
MCT 8 mRNA

10 appears to be a suppressed TRH expression, and therefore

1.5
8 reduced thyroid hormone production, the question could
1.0 6 be addressed by assessing the effect of TRH treatment.
4 When patients were given a TRH infusion, plasma T3 and
0.5 T4 could be increased, but rT3 concentrations also rose
2
0.0 0 (41). However, when TRH was combined with a GH-
secretagogue, this rise in rT3 was prevented, explained by
* * a GH-mediated effect on the inactivating D3 (42). This
3.0 3.0
MCT8 mRNA

treatment also induced an anabolic response, which sug-

gested a causal relationship between low thyroid hormone
2.0 2.0
levels and the impaired anabolism during prolonged crit-
ical illness (40). Furthermore, the negative feedback ex-
RABBIT MODEL

1.0 1.0
erted by thyroid hormones on the thyrotropes was found
to be maintained during TRH infusion, a self-limitation
0.0 0.0
that precludes overstimulation of the thyroid axis (41, 43).
* *
3.0 3.0 Diagnostic implications
MCT10 mRNA

Given the nature of the changes within the thyroid axis

2.0 * 2.0
* roid disease during critical illness is very difficult. Patients
1.0 1.0
0.0 0.0
Figure 2. The top panel represents the circulating thyroid hormone
parameters in acutely stressed (light gray bars, n ⫽ 22) and chronically
ill (dark gray bars, n ⫽ 64) patients. The white bars designate the
normal ranges. The second panel shows the relative MCT8 mRNA
expression levels measured in liver and skeletal muscle of acutely
stressed (light gray) and chronically ill (dark gray) patients. The lower
panels represents the relative expression levels of MCT8 and MCT10 in
liver and muscle of healthy control rabbits (white bars), saline-treated
prolonged ill rabbits (dark gray bars), and T3⫹T4 treated (black bars) ill
rabbits. Data are expressed as mean ⫾ SEM. *, P ⬍ .05 vs acute
values. [Figure was drafted from original data in L. Mebis et al:
Expression of thyroid hormone transporters during critical illness. Eur J
Endocrinol. 2009;161:243–250 (36), with permission. © European
Society of Endocrinology.]
associated with less sepsis susceptibility (38). At the level
of the thyroid hormone receptor (TR), an inverse corre-
lation was observed between the active TR-1/inactive
TR-2 ratio, a surrogate marker of thyroid hormone sen-
sitivity, and the ratio of T3/rT3 in liver biopsies of pro-
evoked by critical illness, the diagnosis of pre-existing thy-
with pre-existing primary hypothyroidism are expected to
reveal low serum levels of T4 and T3 in combination with
high TSH concentrations. However, when primary hypo-
thyroidism and severe nonthyroidal critical illness coin-
cide, TSH levels may be lower than anticipated. Moreover,
serum TSH may be paradoxically low because of iatro-
genic factors such as iodine wound dressings, iodine-con-
taining contrast agents, and drugs such as high-dose cor-
ticosteroids, dopamine, somatostatin, and amiodarone
(30, 44). So, a normal or low TSH during critical illness
does not exclude primary hypothyroidism. Also, the low
T4 and T3 levels in patients with severe hypothyroidism
can be indistinguishable from those values observed in
prolonged nonthyroidal critical illness. A high ratio of
T3/T4 in serum, a low thyroid hormone-binding ratio, and
a low serum rT3 may favor the presence of primary hy-
pothyroidism. However, the diagnostic accuracy is lim-
ited. In these patients, history, physical examination, and
the possible presence of thyroid autoantibodies may give
further clues to the presence of thyroid disease. Repeated
doi: 10.1210/jc.2013-4115
thyroid function tests after improvement of the nonthy-
roidal illness are required to confirm the diagnosis.
Elevated plasma T4 and T3 concentrations are so un-
usual during critical illness that they should always raise
concern of pre-existing hyperthyroidism. However, unde-
tectable TSH has no diagnostic value for hyperthyroidism
during critical illness.
Therapeutic implications
Because the available evidence now indicates that the
acute “low T3 syndrome” appears to be an adaptive re-
sponse partially explained by fasting, treatment is likely
not indicated (22, 23). In contrast, the low T4 and T3 levels
during the prolonged phase of critical illness could be mal-
adaptive. Experimental studies showed that in animal
models of prolonged critical illness and in prolonged crit-
ically ill patients who are receiving nutrition, the syn-
drome can be reversed via hypothalamic-releasing factors,
with an anabolic response at the tissue level (40). How-
ever, the effect on clinical outcome of such a treatment
remains to be investigated, so therapeutic implications are
currently lacking.
An alternative option for treatment could be the ad-
ministration of thyroid hormones T4 or T3 or the combi-
nation to normalize the plasma concentrations. In animal
studies, substitution doses of T4, T3, or their combination
were unable to alter circulating levels of thyroid hor-
mones, likely explained by the increased metabolism of
thyroid hormones during critical illness, perhaps in part
mediated by sulfo-conjugation as was also shown in pa-
tients (45– 48). Three times the substitution dose of T4
normalized plasma T3 concentrations in this model but
resulted in supranormal T4 levels and a rise in rT3. A dose
of T3 that was able to normalize the plasma T3 concen-
trations, five times the substitution dose, suppressed TSH
and T4 to subnormal levels via negative feedback inhibi-
tion. A combination of these doses of T4 and T3 resulted
in dramatic overtreatment. Similar dosing issues were
present in the few available small randomized studies in
critically ill patients, which also did not show outcome
benefits (49 –52).
When and how to treat primary hypothyroidism during
critical illness also remains controversial . When patients
were receiving active treatment for hypothyroidism before
critical illness, it seems wise to continue their usual dose of
thyroid hormone. For myxedema coma, it is generally ac-
cepted that patients should be treated with parenteral in-
fusion of thyroid hormones. However, the proper initia-
tion of replacement therapy during other types of critical
illnesses remains controversial. There is no consensus on
the type of thyroid hormone or on the optimal initial dose
for replacement therapy. Many clinicians prefer a high iv
jcem.endojournals.org 1573
loading dose of 300 –500 ␮g of T4 to quickly reach 50%
of the euthyroid value of T4 (53–55), followed by 50 –100
␮g of iv T4 daily until oral medication can be given. Some
authors have suggested the use of a co-infusion of the
biologically active form of T3 and T4. Escobar-Morreale et
al (56) showed in an animal study that T4 alone did not
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
ensure euthyroidism in all tissues, which was achieved by
combined treatment with T4 and T3. An experimental pro-
tocol for thyroid hormone therapy during prolonged in-
tensive care of presumed hypothyroidism advises admin-
istering a 100- to 200-␮g bolus of T4 iv per 24 hours alone
or, when required to also increase plasma T3, combined with
T3 at 0.6 ␮g/kg ideal body weight per 24 hours in a contin-
uous iv infusion, targeting serum thyroid hormone levels in
the low-normal range (57). When the patients start to re-
cover, a prompt tapering of this dose may be required.
The treatment for primary hyperthyroidism is less af-
fected by concomitant critical illness, except that treat-
ment requirements could be lower in the presence of in-
creased thyroid hormone metabolism. Furthermore, when
patients are receiving active treatment for hyperthyroid-
ism, they should be monitored because of potential tox-
icity of the medication and the impact of other frequently
used medication on thyroid hormone levels.
Hypothalamic-Pituitary-Adrenal Axis
Responses within the hypothalamic-pituitary-
adrenal axis during acute and prolonged critical
illness
The stress hormone cortisol is an essential component
of the fight-or-flight reaction to the stress of illness and
trauma, and both very high and low cortisol levels have
been associated with the risk of death in such patients (58).
Whenever the brain senses a stressful event, activation of
the hypothalamic-pituitary-adrenal axis initiates the re-
lease of the CRH and arginine vasopressin from the hy-
pothalamus, which stimulates the anterior pituitary cor-
ticotrophs to secrete ACTH. High cortisol levels during
critical illness likely contribute to the provision of extra
energy to vital organs by acutely shifting carbohydrate,
fat, and protein metabolism and by delaying anabolism.
Moreover, cortisol likely affects the hemodynamic system
by intravascular fluid retention and by enhancing inotro-
pic and vasopressor responses, respectively, to cat-
echolamines and angiotensin II. In addition, the anti-in-
flammatory effects of cortisol can be interpreted as an
attempt to prevent overactivation of the inflammatory
cascade (59, 60).
During critical illness, plasma cortisol concentrations
are substantially elevated, which is traditionally explained
1574 Boonen and Van den Berghe Novel Insights on Endocrine Changes in the ICU J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582

by severalfold elevated cortisol production in the adrenal work that used a state of the art cortisol-tracer technique
cortex driven by ACTH. However, Vermes et al (61) re- showed that daytime cortisol production during critical
ported only transiently elevated ACTH concentrations in illness was only slightly higher than in healthy subjects.
patients with multiple trauma or sepsis, whereas cortisol Furthermore, cortisol production was only increased in
concentrations remained high. This was recently con- patients with excessive inflammation, whereas it was un-
firmed in a more heterogeneous critically ill patient pop- altered in other critically ill patients (Figure 3) (62). Cor-

Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
ulation. In this study (62), plasma ACTH concentrations tisol breakdown on the other hand was substantially
were found to be suppressed already from ICU admission reduced, irrespective of the inflammatory status, attribut-
onward and stayed below the lower limit of normality able to suppressed expression and activity of A-ring re-
throughout the first week of critical illness. It remains un- ductases in the liver and by suppressed activity of 11␤-
known whether the expected initial ACTH rise in response hydroxysteroid dehydrogenase type 2 in kidney (62). It
to stress was missed in this study and had already occurred remains unclear, however, what is driving the suppression
before ICU admission, for example, in the operating room of these enzymes, but an inverse correlation between ele-
or emergency department. vated plasma concentrations of bile acids and the expres-
Low plasma ACTH in the presence of high plasma cor- sion level of the A-ring reductases could point to bile acids
tisol concentrations has been interpreted as non-ACTH- playing a role (Figure 3) (62, 64). Indeed, bile acids are
driven cortisol production, among which cytokines could potent inhibitors of the cortisol-metabolizing enzymes,
play a role (61, 63). Alternatively, this constellation could both via competitive inhibition and by suppression of gene
be caused by reduced cortisol breakdown suppressing the and protein expression (65– 67).
production of adrenocortical hormones via feedback in- The concept of increasing the bioavailability of cortisol
hibition. In fact, direct evidence of increased cortisol pro- levels primarily in tissues that produce these enzymes, and
duction during critical illness has been lacking. Recent to a lesser extent in the circulation, could be interpreted as

A B
Cortisol Production (mg/h)

60 mg / day P=0.01 P=0.03

of D4-cortisol (liter/min)

4
.6
Plasma Clearance

3.5
3 P=0.34
2.5 .4
30 mg / day 2
1.5
.2
1
.5
0 0
Controls No SIRS SIRS Controls Patients

C P<0.001 D P<0.001 E
1.6 5 P<0.001
5β-reductase protein (10log)

R²=0.36
5β–reductase protein

4 1
5β-reductase mRNA

1.2

3 0
.8
2
-1
.4
1
-2
0 0
Controls Patients Controls Patients -0.25 0.25 0.75 1.25 1.75 2.25
Total Bile Acids (μmol/liter) (10log)
Figure 3. A, Cortisol production in critically ill patients with the SIRS (n ⫽ 7; dark gray bar) and no SIRS (n ⫽ 4; light gray bar) compared to
controls (n ⫽ 9; white bar). Based on these results, 24-hour cortisol production was estimated and depicted with arrows. B, Cortisol plasma
clearance as assessed with a small dose of deuterated-cortisol tracer. Bar charts represent means and SE values. C–E, mRNA and protein expression
of 5␤-reductase in liver of 20 controls (white bars) and 44 patients (gray bars) and the relation to plasma total bile acid concentrations. Bar charts
represent means and SE values. The mRNA data are expressed, normalized to glyceraldehyde-3-phosphate dehydrogenase, as a fold difference
from the mean of the controls. Protein data are expressed, normalized for CK-18 protein expression, as a fold difference from the mean of the
controls. [Figure was drafted from original data from E. Boonen et al: Reduced cortisol metabolism during critical illness. N Engl J Med. 2013;368:
1477–1488 (62), with permission. © Massachusetts Medical Society.]
doi: 10.1210/jc.2013-4115
a highly economic way to keep cortisol levels high without
spending too much energy producing it. This concept is
further supported by low plasma cortisol binding globulin
levels in critical illness, causing increased levels of free
cortisol, the biologically active form. Furthermore, as
such, cortisol is elevated locally in liver and kidney, where
it is needed for an optimal fight-or-flight response, with-
out an undue exposure of immune cells and vulnerable
target tissues such as skeletal muscle or brain to the del-
eterious side effects of hypercortisolism. The local effects
of cortisol appear to be further regulated at the level of
glucocorticoid receptor (GR) expression. Previous work
indeed showed suppressed expression of GR in white
blood cells of critically ill children, which could be a way
to allow the innate immune response to effectively protect
the host against infections in the presence of hypercorti-
solism (68). Clearly, this novel concept of tissue-specific
regulation of glucocorticoid activity during critical illness
requires further investigation.
The new insight that during critical illness cortisol me-
tabolism is suppressed, contributing to hypercortisolism,
could theoretically explain the concomitantly low plasma
ACTH concentrations via negative feedback inhibition at
the level of the pituitary gland and/or the hypothalamus,
but studies assessing this at the tissue level are currently
lacking. It remains unclear whether such a sustained sup-
pressed ACTH secretion could cause adrenal atrophy in
the prolonged phase of critical illness. However, this could
explain the reported 20-fold higher incidence of symp-
tomatic adrenal insufficiency in critically ill patients being
treated in the ICU for more than 14 days (69). Other fac-
tors contributing to adrenal failure are also possible, such
as endothelial dysfunction (70, 71), although conforma-
tional human studies are lacking.
Diagnostic implications
Since the last decade, reference is made to “relative
adrenal insufficiency” in the context of critical illness (72–
74). It refers to the condition in which, despite a maximally
ACTH-activated adrenal cortex in response to critical ill-
ness, the cortisol production is still insufficient to generate
enough GR and mineralocorticoid receptor activation to
maintain hemodynamic stability. From large association
studies, such a condition is thought to be identifiable by an
insufficient rise (⬍9 ␮g/dL) in plasma cortisol in response
to a 250-␮g ACTH bolus, irrespective of the baseline
plasma cortisol concentration, which is usually much
higher than in healthy humans (72). In such a condition of
insufficiently increased cortisol production, a very high
plasma ACTH concentration would be expected. How-
ever, the recent robust findings that ACTH plasma con-
centrations are suppressed, that cortisol production is not
jcem.endojournals.org 1575
much elevated, if at all, and that instead reduced cortisol
breakdown plays a major role during critical illness, fur-
ther complicate the issue of diagnostic criteria for adrenal
failure in that setting. Moreover, it was recently shown
that cortisol responses to ACTH stimulation in critically ill
patients correlated positively with both cortisol produc-
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
tion rate and cortisol plasma clearance, but patients who
revealed the lowest response to ACTH, to the extent of
absolute adrenal failure, were the ones with the most sup-
pressed cortisol breakdown, whereas their cortisol pro-
duction was similar to healthy subjects (65). These find-
ings hint that a low cortisol response to an ACTH injection
reflects the degree of negative feedback inhibition exerted
by the high levels of circulating cortisol, a situation similar
to patients treated with exogenous glucocorticoids for an
extended time, who also reveal a suppressed response to
ACTH injection. Whether this low response during critical
illness indicates that cortisol availability would be “insuf-
ficient” to cope with the stress of illness remains unclear.
Alternatively, a random total cortisol of ⬍ 10 ␮g/dL
during critical illness has been suggested for the diagnosis
of “relative adrenal insufficiency” (75). However, total
plasma cortisol concentration is the net effect of adrenal
production and secretion, distribution, binding, and elim-
ination of cortisol. Judging the adequacy of the adrenal
cortisol production in response to critical illness based on
a single measurement of total plasma cortisol is merely
indicative. Furthermore, circulating total cortisol concen-
trations do not reveal the glucocorticoid effect. During
crucial illness suppressed circulating levels of the binding
proteins, cortisol binding globulin (CBG) and albumin, as
well as decreased CBG binding affinity via increased cleav-
age from CBG at inflammatory loci or by increased tem-
perature were established (76 –79). Since only free cortisol
can pass the cell membrane to bind to GR and plasma, free
cortisol may be more appropriate to assess HPA-axis func-
tion. However, more research is needed because plasma
free cortisol assays are not readily available, and normal
ranges for plasma free cortisol during critical illness have
not been defined. Additionally, increasing evidence from
both animal and human experiments suggests altered GR
regulation during critical illness (68, 80 – 84), precluding
conclusions about “adequacy” of cortisol availability and
function during illness. Finally, assays to quantify plasma
cortisol concentrations are often inaccurate and vary sub-
stantially (85), making it impossible to identify one cutoff
value for clinical practice.
Recently, measuring interstitial cortisol levels was in-
troduced to assess the amount of active tissue cortisol lev-
els in critically ill patients (86, 87). Therefore, a microdi-
alysis catheter is inserted into the sc adipose tissue.
However, critical illness presents frequently with edema,
1576 Boonen and Van den Berghe Novel Insights on Endocrine Changes in the ICU J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582
and regional blood flow is variable. Furthermore, the sc
adipose tissue is not the main target tissue for cortisol, nor
is it the main cortisol-metabolizing organ during critical
illness (62).
Therapeutic implications
It is generally accepted that patients with an established
diagnosis of primary or central adrenal failure or patients
on chronic treatment with systemic glucocorticoids before
critical illness should receive additional coverage to cope
with the acute stress (53, 88). Also, patients who are di-
agnosed with an acute Addisonian crisis in the ICU are
typically treated with high doses of glucocorticoids. This
therapeutic strategy is based on the assumption that cor-
tisol production is increased severalfold in critical illness.
The conventional treatment proposes the administration
of a bolus of 100-mg hydrocortisone followed by 50 to
100 mg every 6 hours on the first day, 50 mg every 6 hours
on the second day, and 25 mg every 6 hours on the third
day, tapering to a maintenance dose by the fourth to fifth
day (53, 88).
The dose of hydrocortisone advised for treatment of
“relative adrenal failure” is another controversial issue.
The proposed dose of 300 mg of hydrocortisone per day,
referred to as “low dose” in the literature, is in fact ap-
proximately 10 times higher than the normal amount of
daily cortisol production in healthy humans (89 –92) and
between 3- and 6-fold higher than the production that
now has been quantified in critically ill patients (Figure 3).
In view of the substantially reduced cortisol breakdown
during critical illness, the currently proposed doses for
adrenal failure during critical illness may be too high. This
may further explain why the multicenter randomized con-
trolled study that assessed the effect of hydrocortisone
treatment could not confirm the benefit that was originally
observed in the pioneer trial (89, 92).
Also, the duration of treatment is under debate. Treat-
ing critically ill patients with glucocorticoids in too high of
a dose for too long of a period could inferentially aggra-
vate the loss of lean tissue, increase the risk of myopathy,
and prolong the ICU dependence, which could increase the
susceptibility to potentially lethal complications (93, 94).
Finally, because glucocorticoid sensitivity likely varies
among individuals (95) and among cell types in critically
ill patients (68, 81, 83) and glucocorticoid treatment may
down-regulate GR-␣ via induction of miR-124, the dosing
issue is further complicated (96). Moreover, single nucle-
otide polymorphisms in the GR gene, with an altered re-
sponse to glucocorticoids, have been identified (97). How-
ever, it remains a challenge to identify specific clinical
biomarkers of GR activation to guide optimal glucocor-
ticoid therapy for individual patients and illnesses.
Based on the results of stable isotope studies (62), a dose
of ⫾ 60 mg of hydrocortisone, equivalent to about a dou-
bling of the normal daily cortisol production, may be in-
teresting to investigate further when patients at risk can be
identified. A fast tapering down to the lowest effective
dose should limit the adverse effects of excessive amounts
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
of glucocorticoids during critical illness.
The Hyperglycemic Response to Critical
Illness: To Treat or Not to Treat?
Blood glucose and critical illness: robust
associative data
In humans, the natural endocrine and immunological
responses to stress ensure adequate availability of glucose
by activating gluconeogenesis and by reducing the sensi-
tivity to insulin for those organs and tissues that predom-
inantly rely upon glucose as metabolic substrate, such as
the brain and blood cells. In young and lean patients not
receiving macronutrients, this stress response will main-
tain normoglycemia. However, when patients are older,
are overweight, suffer from chronic comorbidity, or re-
ceive drugs that affect insulin sensitivity or enteral/paren-
teral nutrition, the circulating glucose concentrations usu-
ally rise quickly above the upper limit of normality (98 –
102), which could be adaptive or maladaptive. In the
condition of prolonged critical illness, stress-induced hy-
perglycemia may be quite severe and may persist for a long
period of time. Hyperglycemia in critically ill patients has
repeatedly been shown to be associated with a risk of mor-
tality, an association that appears to have a J-shape with
the lowest risk in the normoglycemic zone (Figure 4) (103).
In critically ill patients with established diabetes mellitus,
the J-shaped curve is significantly blunted in the hyper-
glycemic zone, and the nadir is shifted to higher blood
glucose levels (103, 104).
Hyperglycemia and adverse outcome: cause or
consequence?
The first randomized controlled trial on blood glucose
management was the 2001 Leuven Surgical ICU study
(105). In this study, a “strictly normal level for fasting
blood glucose,” ie, 80 –110 mg/dL, was targeted in the
intervention group, as compared to the “usual care” of
adult surgical ICU patients in the year 2000, which was to
tolerate hyperglycemia up to 215 mg/dL. The study was
highly standardized, resulting in a strong internal validity.
For example, frequent blood glucose measurements (in-
terval, 0.5– 4 h) on whole arterial blood by an accurate
blood gas analyzer were done by well-trained nurse, and
insulin was continuously infused exclusively via a dedi-
doi: 10.1210/jc.2013-4115 jcem.endojournals.org 1577

DIFFERENCES IN DESIGN EXPECTED OUTCOMES BASED ON LEUVEN TRIALS

The Leuven comparison Cumulative risk in-hospital mortality

“Don’t touch”
The NICE-SUGAR
MORTALITY

comparison .3

Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
.2

.1

0 100 200 300 400 500 600

hypo normal for age “renal threshold”
BLOOD GLUCOSE days
Figure 4. Different designs of key intervention trials and expected outcome benefits. The left panel shows J-shaped association curve between
blood glucose and risk of death. The NICE-SUGAR trial was executed in the flatter part of the J-shaped curve. A very small benefit from aiming at
lowering blood glucose further down from an intermediate level to strict normoglycemia was hereby traded off against a similar risk of harm by
hypoglycemia, particularly when using inaccurate tools. The right panel shows the dose response in the two adult Leuven trials compared to the
NICE-SUGAR trial. Black circles represent blood glucose ⬎ 150mg/dl, dark grey circles represent blood glucose 110 –150mg/dl and light grey circles
represent blood glucose ⬍ 110mg/dl. The maximal benefit that could be expected from lowering blood glucose from an intermediate level to
normoglycemia is ⬍ 1%, provided blood glucose could be perfectly separated between the two study arms. To confidently conclude that such a
small benefit is not present, 70 000 patients should have been included. Hence, NICE-SUGAR, with 6100 patients, was in fact underpowered to
address this hypothesis. [Reproduced from G. Van den Berghe: Intensive insulin therapy in the ICU–reconciling the evidence. Nat Rev Endocrinol.
2012;8:374 –378 (127), with permission. © Macmillan Publishers Limited.]

cated lumen of a central venous line with an accurate sy-
ringe pump. Maintaining strict normoglycemia lowered
ICU and in-hospital mortality and reduced morbidity by
preventing organ failure, reflected in a shorter duration of
mechanical ventilation, a decreased incidence of acute kid-
ney failure, severe infections, and critical illness polyneu-
ropathy. In a second study performed in patients admitted
to a medical ICU in Leuven, these morbidity benefits were
confirmed (106). A subsequent randomized controlled
study was performed in critically ill children, in which the
intervention group was targeted to normal fasting glucose
levels for the age groups (50 – 80 mg/dL for infants, and
70 –100 mg/dL for children) as compared with tolerating
hyperglycemia up to 215 mg/dL (107). Also in this young
patient population, the intervention reduced ICU morbid-
ity and mortality and also had long-term beneficial effects
on neurocognitive development up to 4 years after inclu-
sion in the study (107, 108). In a subsequent study, tar-
geting the much higher adult range for normal fasting
blood glucose levels in such young infants in the ICU did
not alter the level of blood glucose concentration or the
outcome (109), suggesting that the normal fasting level is
key to preventing toxicity of hyperglycemia in each age
group. The underlying mechanisms of hyperglycemia-in-
duced toxicity were identified to involve cellular damage
occurring in those cells that do not require insulin for
glucose uptake, such as hepatocytes, renal tubular cells,
the endothelium, immune cells, and neurons (93, 110 –
113). Soon after the first Leuven study was published, the
intervention was swiftly implemented in clinical practice
worldwide (114, 115). After several smaller studies, the
NICE-SUGAR (Normoglycemia in Intensive Care Evalu-
ation and Survival Using Glucose Algorithm Regulation)
multicenter trial was designed to be the definitive study to
answer this question. The study compared tight blood glu-
cose control to a normoglycemic target (80 –100 mg/dL) in
the intervention group with an intermediate target of 140 –
180 mg/dL in the control group (116). The study revealed
that blood glucose control to a normoglycemic target in-
creased mortality as compared with the intermediate level
in the control group (116), subsequently explained by a
13-fold increase in hypoglycemia (117). Because this study
was designed for a high external validity, the first conclu-
sion is that very tight blood glucose control is not readily
applicable in general daily clinical practice. However, the
usual care had already evolved significantly between the
first Leuven study and the start of NICE-SUGAR; toler-
ating excessive hyperglycemia was now the new no-go
zone, compared to the 215 mg/dL tolerance threshold 5
years earlier. Second, due to its pragmatic nature, there
was no emphasis on standardization in NICE-SUGAR. All
sorts of glucose measurement methodologies were al-
lowed, and practitioners were not specifically trained to
perform the complex treatment. Now, it has become clear
that tight blood glucose control requires accurate blood
gas analyzers, like those used in the Leuven studies, to
target a narrow range of blood glucose (118). It is also
clear that extensive experience is crucial to avoid unde-
tected episodes of hypoglycemia and to treat hypoglyce-
mia when it occurs. Certainly profound, prolonged/unde-
1578 Boonen and Van den Berghe Novel Insights on Endocrine Changes in the ICU J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582
tected hypoglycemia can have grave consequences and
may even result in death. Hence, hypoglycemia should be
avoided as much as possible. Nevertheless, recent data
show that in cardiac patients and in critically ill children,
iatrogenic hypoglycemia may not by itself affect outcome
(108, 119, 120). Spontaneous hypoglycemia is in contrast
a strong predictor of poor outcome. For example, patients
with liver failure, acute kidney injury requiring renal re-
placement therapy, diabetes mellitus, and septic shock
have higher risk to develop spontaneous hypoglycemia.
Furthermore, adequate treatment of hypoglycemia is es-
sential to avoid rebound hyperglycemia, which causes
brain damage (121). Detailed protocols for prompt and
gentle correction of hypoglycemia are often not in place,
which again contrasts with the Leuven studies (116).
Therapeutic implications: how to translate this
into general clinical practice?
What then could be a sensible approach for daily prac-
tice? Tight blood glucose control with current technolo-
gies and experience is not yet ready to be broadly imple-
mented in every ICU, as clearly demonstrated by NICE-
SUGAR. Post hoc analyses of the Leuven clinical trials
revealed that the bulk of the beneficial effects of blood
glucose control lay in bringing overt hyperglycemia to
moderate levels (Figure 4) (122, 123). More can be gained
by further tightening the glycemic control, but it requires
a substantial investment in training and technology to do
this safely. Hence, targeting blood glucose below 145
mg/dL seems a reasonable compromise (115). Critically ill
diabetic patients may benefit from treatment to somewhat
higher glycemic targets, depending on their premorbid lev-
els (102–103). However, irrespective of the chosen target
level, several methodological aspects ought to be taken
into account to assure patient safety whenever insulin
treatment is used. These include frequent blood glucose
measurements, the use of on-site blood gas analyzers as the
preferred measurement tool, the avoidance of capillary
blood samples, and the continuous infusion of insulin with
accurate syringe pumps through a dedicated lumen of a
central venous catheter. Finally, insulin dosing decisions
should not be based on a sliding scale system, but instead
on a (computerized) algorithm that was clinically vali-
dated for critically ill patients (124).
Conclusions
Recent studies generated important novel insights in the
endocrine and metabolic responses to critical illness. Al-
though many aspects remain unresolved, an important
recent insight with therapeutic implications is that most of
the acute endocrine responses are likely adaptive and thus
should probably not be treated. Nevertheless, many pa-
tients who survived the initial phase of critical illness still
remain in the ICU for long periods and face a risk of death
that increases steadily with every day that recovery does
not set in. Hence, more work is required to find better
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
treatments to further prevent protracted critical illness, to
enhance recovery from organ failure, and to optimize
rehabilitation.
Novel Insights in Endocrine Changes in
Critical Illness
• Part of the acute fall in T3 plasma concentrations during
critical illness is related to the concomitant fasting, and
this part of the response seems adaptive.
• Cortisol production is only moderately increased dur-
ing critical illness and is only increased in patients suf-
fering from the systemic inflammatory response syn-
drome (SIRS). Cortisol production is unaltered in
patients without SIRS, in the face of severalfold higher
plasma cortisol in all patients.
• Cortisol plasma clearance is substantially reduced in all
critically ill patients and contributes substantially to
hypercortisolism during critical illness, irrespective of
the type and severity of illness and irrespective of the
inflammation status.
• The largest benefit of blood glucose control may be
brought about by preventing overt hyperglycemia;
hence, targeting blood glucose to intermediate ranges
during critical illness seems a reasonable compromise.
Acknowledgments
Address all correspondence and requests for reprints to: Greet
Van den Berghe, Clinical Division and Laboratory of Intensive
Care Medicine, KU Leuven, Herestraat 49, B-3000 Leuven, Bel-
gium. E-mail: greet.vandenberghe@med.kuleuven.be.
This work was supported by research grants from the Fund
for Scientific Research Flanders Belgium, by the Methusalem
Program funded by the Flemish Government, and by the Euro-
pean Research Council under the European Union’s Seventh
Framework Program (FP7/2007–2013 ERC Advanced Grant
Agreement no. 307523).
Disclosure Summary: The authors have no conflict of interest
to declare.
References
1. Casaer MP, Langouche L, Coudyzer W, et al. Impact of early par-
enteral nutrition on muscle and adipose tissue compartments during
critical illness. Crit Care Med. 2013;41:2298 –2309.
doi: 10.1210/jc.2013-4115
2. Puthucheary ZA, Rawal J, McPhail M, et al. Acute skeletal muscle
wasting in critical illness. JAMA. 2013;310:1591–1600.
3. Villet S, Chiolero RL, Bollmann MD, et al. Negative impact of hy-
pocaloric feeding and energy balance on clinical outcome in ICU
patients. Clin Nutr. 2005;24:502–509.
4. Hermans G, Casaer MP, Clerckx B, et al. Effect of tolerating ma-
cronutrient deficit on the development of intensive-care unit ac-
quired weakness: a subanalysis of the EPaNIC trial. Lancet Respir
Med. 2013;1:621– 629.
5. Vanhorebeek I, Gunst J, Derde S, et al. Insufficient activation of
autophagy allows cellular damage to accumulate in critically ill pa-
tients. J Clin Endocrinol Metab. 2011;96:E633–E645.
6. Chopra IJ, Huang TS, Beredo A, Solomon DH, Chua Teco GN,
Mead JF. Evidence for an inhibitor of extrathyroidal conversion of
thyroxine to 3,5,3⬘-triiodothyronine in sera of patients with non-
thyroidal illnesses. J Clin Endocrinol Metab. 1985;60:666 – 672.
7. Michalaki M, Vagenakis AG, Makri M, Kalfarentzos F, Kyriazo-
poulou V. Dissociation of the early decline in serum T(3) concen-
tration and serum IL-6 rise and TNF␣ in nonthyroidal illness syn-
drome induced by abdominal surgery. J Clin Endocrinol Metab.
2001;86:4198 – 4205.
8. Bartalena L, Martino E, Brandi LS, et al. Lack of nocturnal serum
thyrotropin surge after surgery. J Clin Endocrinol Metab. 1990;70:
293–296.
9. Romijn JA, Wiersinga WM. Decreased nocturnal surge of thyrotro-
pin in nonthyroidal illness. J Clin Endocrinol Metab. 1990;70:35–
42.
10. Redout EM, van der Toorn A, Zuidwijk MJ, et al. Antioxidant
treatment attenuates pulmonary arterial hypertension-induced
heart failure. Am J Physiol Heart Circ Physiol. 2010;298:H1038 –
H1047.
11. Wajner SM, Goemann IM, Bueno AL, Larsen PR, Maia AL. IL-6
promotes nonthyroidal illness syndrome by blocking thyroxine ac-
tivation while promoting thyroid hormone inactivation in human
cells. J Clin Invest. 2011;121:1834 –1845.
12. Boelen A, Kwakkel J, Fliers E. Beyond low plasma T3: local thyroid
hormone metabolism during inflammation and infection. Endocr
Rev. 2011;32:670 – 693.
13. van der Poll T, Van Zee KJ, Endert E, et al. Interleukin-1 receptor
blockade does not affect endotoxin-induced changes in plasma thy-
roid hormone and thyrotropin concentrations in man. J Clin Endo-
crinol Metab. 1995;80:1341–1346.
14. Lim CF, Docter R, Visser TJ, et al. Inhibition of thyroxine transport
into cultured rat hepatocytes by serum of nonuremic critically ill
patients: effects of bilirubin and nonesterified fatty acids. J Clin
Endocrinol Metab. 1993;76:1165–1172.
15. Gardner DF, Kaplan MM, Stanley CA, Utiger RD. Effect of tri-
iodothyronine replacement on the metabolic and pituitary responses
to starvation. N Engl J Med. 1979;300:579 –584.
16. Moshang T Jr, Parks JS, Baker L, et al. Low serum triiodothyronine
in patients with anorexia nervosa. J Clin Endocrinol Metab. 1975;
40:470 – 473.
17. Peeters RP, Wouters PJ, van Toor H, Kaptein E, Visser TJ, Van den
Berghe G. Serum 3,3⬘,5⬘-triiodothyronine (rT3) and 3,5,3⬘-triiodo-
thyronine/rT3 are prognostic markers in critically ill patients and are
associated with postmortem tissue deiodinase activities. J Clin En-
docrinol Metab. 2005;90:4559 – 4565.
18. Rothwell PM, Lawler PG. Prediction of outcome in intensive care
patients using endocrine parameters. Crit Care Med. 1995;23:78 –
83.
19. Klemperer JD, Klein I, Gomez M, et al. Thyroid hormone treatment
after coronary-artery bypass surgery. N Engl J Med. 1995;333:
1522–1527.
20. Mullis-Jansson SL, Argenziano M, Corwin S, et al. A randomized
double-blind study of the effect of triiodothyronine on cardiac func-
tion and morbidity after coronary bypass surgery. J Thorac Cardio-
vasc Surg. 1999;117:1128 –1134.
21. Casaer MP, Mesotten D, Hermans G, et al. Early versus late par-
jcem.endojournals.org 1579
enteral nutrition in critically ill adults. N Engl J Med. 2011;365:
506 –517.
22. Langouche L, Vander Perre S, Marques M, et al. Impact of early
nutrient restriction during critical illness on the nonthyroidal illness
syndrome and its relation with outcome: a randomized, controlled
clinical study. J Clin Endocrinol Metab. 2013;98:1006 –1013.
23. Mebis L, Eerdekens A, Güiza F, et al. Contribution of nutritional
deficit to the pathogenesis of the nonthyroidal illness syndrome in
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
critical illness: a rabbit model study. Endocrinology. 2012;153:973–
984.
24. Boelen A, Boorsma J, Kwakkel J, et al. Type 3 deiodinase is highly
expressed in infiltrating neutrophilic granulocytes in response to
acute bacterial infection. Thyroid. 2008;18:1095–1103.
25. Van den Berghe G, de Zegher F, Veldhuis JD, et al. Thyrotrophin
and prolactin release in prolonged critical illness: dynamics of spon-
taneous secretion and effects of growth hormone-secretagogues.
Clin Endocrinol (Oxf). 1997;47:599 – 612.
26. Fliers E, Guldenaar SE, Wiersinga WM, Swaab DF. Decreased hy-
pothalamic thyrotropin-releasing hormone gene expression in pa-
tients with nonthyroidal illness. J Clin Endocrinol Metab. 1997;82:
4032– 4036.
27. Bacci V, Schussler GC, Kaplan TB. The relationship between serum
triiodothyronine and thyrotropin during systemic illness. J Clin En-
docrinol Metab. 1982;54:1229 –1235.
28. Damas P, Reuter A, Gysen P, Demonty J, Lamy M, Franchimont P.
Tumor necrosis factor and interleukin-1 serum levels during severe
sepsis in humans. Crit Care Med. 1989;17:975–978.
29. Faglia G, Ferrari C, Beck-Peccoz P, Spada A, Travaglini P, Ambrosi
B. Reduced plasma thyrotropin response to thyrotropin releasing
hormone after dexamethasone administration in normal subjects.
Horm Metab Res. 1973;5:289 –292.
30. Van den Berghe G, de Zegher F, Lauwers P. Dopamine and the sick
euthyroid syndrome in critical illness. Clin Endocrinol (Oxf). 1994;
41:731–737.
31. Van den Berghe G, de Zegher F, Lauwers P. Dopamine suppresses
pituitary function in infants and children. Crit Care Med. 1994;22:
1747–1753.
32. Boelen A, Kwakkel J, Thijssen-Timmer DC, Alkemade A, Fliers E,
Wiersinga WM. Simultaneous changes in central and peripheral
components of the hypothalamus-pituitary-thyroid axis in lipopo-
lysaccharide-induced acute illness in mice. J Endocrinol. 2004;182:
315–323.
33. Mebis L, Debaveye Y, Ellger B, et al. Changes in the central com-
ponent of the hypothalamus-pituitary-thyroid axis in a rabbit model
of prolonged critical illness. Crit Care. 2009;13:R147.
34. Alkemade A, Friesema EC, Kuiper GG, et al. Novel neuroanatomi-
cal pathways for thyroid hormone action in the human anterior
pituitary. Eur J Endocrinol. 2006;154:491–500.
35. Langouche L, Princen L, Gunst J, Güiza F, Derde S, Van den Berghe
G. Anterior pituitary morphology and hormone production during
sustained critical illness in a rabbit model. Horm Metab Res. 2013;
45:277–282.
36. Mebis L, Paletta D, Debaveye Y, et al. Expression of thyroid hor-
mone transporters during critical illness. Eur J Endocrinol. 2009;
161:243–250.
37. Mebis L, Langouche L, Visser TJ, Van den Berghe G. The type II
iodothyronine deiodinase is up-regulated in skeletal muscle during
prolonged critical illness. J Clin Endocrinol Metab. 2007;92:3330 –
3333.
38. Ma SF, Xie L, Pino-Yanes M, et al. Type 2 deiodinase and host
responses of sepsis and acute lung injury. Am J Respir Cell Mol Biol.
2011;45:1203–1211.
39. Thijssen-Timmer DC, Peeters RP, Wouters P, et al. Thyroid hor-
mone receptor isoform expression in livers of critically ill patients.
Thyroid. 2007;17:105–112.
40. Van den Berghe G, Wouters P, Weekers F, et al. Reactivation of
pituitary hormone release and metabolic improvement by infusion
of growth hormone-releasing peptide and thyrotropin-releasing
1580 Boonen and Van den Berghe Novel Insights on Endocrine Changes in the ICU J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582
hormone in patients with protracted critical illness. J Clin Endocri-
nol Metab. 1999;84:1311–1323.
41. Van den Berghe G, de Zegher F, Baxter RC, et al. Neuroendocri-
nology of prolonged critical illness: effects of exogenous thyrotro-
pin-releasing hormone and its combination with growth hormone
secretagogues. J Clin Endocrinol Metab. 1998;83:309 –319.
42. Weekers F, Michalaki M, Coopmans W, et al. Endocrine and met-
abolic effects of growth hormone (GH) compared with GH-releas-
ing peptide, thyrotropin-releasing hormone, and insulin infusion in
a rabbit model of prolonged critical illness. Endocrinology. 2004;
145:205–213.
43. Van den Berghe G, Baxter RC, Weekers F, Wouters P, Bowers CY,
Iranmanesh A, Veldhuis JD, Bouillon R. The combined administra-
tion of GH-releasing peptide-2 (GHRP-2), TRH and GnRH to men
with prolonged critical illness evokes superior endocrine and met-
abolic effects compared to treatment with GHRP-2 alone. Clin En-
docrinol (Oxf). 2002;56:655– 669.
44. Mebis L, van den Berghe G. The hypothalamus-pituitary-thyroid
axis in critical illness. Neth J Med. 2009;67:332–340.
45. Debaveye Y, Ellger B, Mebis L, et al. Tissue deiodinase activity
during prolonged critical illness: effects of exogenous thyrotropin-
releasing hormone and its combination with growth hormone-re-
leasing peptide-2. Endocrinology. 2005;146:5604 –5611.
46. Debaveye Y, Ellger B, Mebis L, Darras VM, Van den Berghe G.
Regulation of tissue iodothyronine deiodinase activity in a model of
prolonged critical illness. Thyroid. 2008;18:551–560.
47. Debaveye Y, Ellger B, Mebis L, Visser TJ, Darras VM, Van den
Berghe G. Effects of substitution and high-dose thyroid hormone
therapy on deiodination, sulfoconjugation, and tissue thyroid hor-
mone levels in prolonged critically ill rabbits. Endocrinology. 2008;
149:4218 – 4228.
48. Peeters RP, van der Geyten S, Wouters PJ, et al. Tissue thyroid
hormone levels in critical illness. J Clin Endocrinol Metab. 2005;
90:6498 – 6507.
49. Acker CG, Singh AR, Flick RP, Bernardini J, Greenberg A, Johnson
JP. A trial of thyroxine in acute renal failure. Kidney Int. 2000;57:
293–298.
50. Becker RA, Vaughan GM, Ziegler MG, et al. Hypermetabolic low
triiodothyronine syndrome of burn injury. Crit Care Med. 1982;
10:870 – 875.
51. Brent GA, Hershman JM. Thyroxine therapy in patients with severe
nonthyroidal illnesses and low serum thyroxine concentration.
J Clin Endocrinol Metab. 1986;63:1– 8.
52. Sirlak M, Yazicioglu L, Inan MB, et al. Oral thyroid hormone pre-
treatment in left ventricular dysfunction. Eur J Cardiothorac Surg.
2004;26:720 –725.
53. Debaveye Y, Vandenbrande J, Van den Berghe G. Endocrine emer-
gencies. In: Tubaro M, Danchin N, Filippatos G, Goldstein P,
Vranckx P, Zahger D, eds. The ESC Textbook of Intensive and
Acute Cardiac Care. New York: Oxford University Press; 2011:
709 –717.
54. Nicoloff JT. Thyroid storm and myxedema coma. Med Clin North
Am. 1985;69:1005–1017.
55. Ringel MD. Management of hypothyroidism and hyperthyroidism
in the intensive care unit. Crit Care Clin. 2001;17:59 –74.
56. Escobar-Morreale HF, Obregón MJ, Escobar del Rey F, Morreale de
Escobar G. Replacement therapy for hypothyroidism with thyrox-
ine alone does not ensure euthyroidism in all tissues, as studied in
thyroidectomized rats. J Clin Invest. 1995;96:2828 –2838.
57. Van den Berghe G. Endocrine aspects of critical care medicine. In:
Jameson JL, De Groot LJ, eds. Endocrinology: Adult and Pediatric.
St Louis, MO: WB Saunders; 2010:2084 –2085.
58. Finlay WE, McKee JI. Serum cortisol levels in severely stressed pa-
tients. Lancet. 1982;1:1414 –1415.
59. Munck A, Guyre PM, Holbrook NJ. Physiological functions of glu-
cocorticoids in stress and their relation to pharmacological actions.
Endocr Rev. 1984;5:25– 44.
60. Starling EH. The wisdom of the body: the Harveian Oration, de-
livered before The Royal College of Physicians of London on St.
Luke’s Day, 1923. Br Med J. 1923;2:685– 690.
61. Vermes I, Beishuizen A, Hampsink RM, Haanen C. Dissociation of
plasma adrenocorticotropin and cortisol levels in critically ill pa-
tients: possible role of endothelin and atrial natriuretic hormone.
J Clin Endocrinol Metab. 1995;80:1238 –1242.
62. Boonen E, Vervenne H, Meersseman P, et al. Reduced cortisol me-
tabolism during critical illness. N Engl J Med. 2013;368:1477–
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
1488.
63. Bornstein SR, Engeland WC, Ehrhart-Bornstein M, Herman JP. Dis-
sociation of ACTH and glucocorticoids. Trends Endocrinol Metab.
2008;19:175–180.
64. Vanwijngaerden YM, Wauters J, Langouche L, et al. Critical illness
evokes elevated circulating bile acids related to altered hepatic trans-
porter and nuclear receptor expression. Hepatology. 2011;54:
1741–1752.
65. Ackermann D, Vogt B, Escher G, et al. Inhibition of 11␤-hydrox-
ysteroid dehydrogenase by bile acids in rats with cirrhosis. Hepa-
tology. 1999;30:623– 629.
66. McNeilly AD, Macfarlane DP, O’Flaherty E, et al. Bile acids mod-
ulate glucocorticoid metabolism and the hypothalamic-pituitary-
adrenal axis in obstructive jaundice. J Hepatol. 2010;52:705–711.
67. Stauffer AT, Rochat MK, Dick B, Frey FJ, Odermatt A. Chenode-
oxycholic acid and deoxycholic acid inhibit 11 ␤-hydroxysteroid
dehydrogenase type 2 and cause cortisol-induced transcriptional
activation of the mineralocorticoid receptor. J Biol Chem. 2002;
277:26286 –26292.
68. van den Akker EL, Koper JW, Joosten K, et al. Glucocorticoid re-
ceptor mRNA levels are selectively decreased in neutrophils of chil-
dren with sepsis. Intensive Care Med. 2009;35:1247–1254.
69. Barquist E, Kirton O. Adrenal insufficiency in the surgical intensive
care unit patient. J Trauma. 1997;42:27–31.
70. Kanczkowski W, Chatzigeorgiou A, Grossklaus S, Sprott D, Born-
stein SR, Chavakis T. Role of the endothelial-derived endogenous
anti-inflammatory factor Del-1 in inflammation-mediated adrenal
gland dysfunction. Endocrinology. 2013;154:1181–1189.
71. Prigent H, Maxime V, Annane D. Science review: mechanisms of
impaired adrenal function in sepsis and molecular actions of gluco-
corticoids. Crit Care. 2004;8:243–252.
72. Annane D, Sébille V, Troché G, Raphaël JC, Gajdos P, Bellissant E.
A 3-level prognostic classification in septic shock based on cortisol
levels and cortisol response to corticotropin. JAMA. 2000;283:
1038 –1045.
73. Beishuizen A, Vermes I, Hylkema BS, Haanen C. Relative eosino-
philia and functional adrenal insufficiency in critically ill patients.
Lancet. 1999;353:1675–1676.
74. Richards ML, Caplan RH, Wickus GG, Lambert PJ, Kisken WA.
The rapid low-dose (1 microgram) cosyntropin test in the immediate
postoperative period: results in elderly subjects after major abdom-
inal surgery. Surgery. 1999;125:431– 440.
75. Marik PE, Pastores SM, Annane D, et al. Recommendations for the
diagnosis and management of corticosteroid insufficiency in criti-
cally ill adult patients: consensus statements from an international
task force by the American College of Critical Care Medicine. Crit
Care Med. 2008;36:1937–1949.
76. Chan WL, Carrell RW, Zhou A, Read RJ. How changes in affinity
of corticosteroid-binding globulin modulate free cortisol concen-
tration. J Clin Endocrinol Metab. 2013;98:3315–3322.
77. Hamrahian AH, Oseni TS, Arafah BM. Measurements of serum free
cortisol in critically ill patients. N Engl J Med. 2004;350:1629 –
1638.
78. Holland PC, Hancock SW, Hodge D, Thompson D, Shires S, Evans
S. Degradation of albumin in meningococcal sepsis. Lancet. 2001;
357:2102–2104.
79. Pugeat M, Bonneton A, Perrot D, et al. Decreased immunoreactivity
and binding activity of corticosteroid-binding globulin in serum in
septic shock. Clin Chem. 1989;35:1675–1679.
80. Bergquist M, Nurkkala M, Rylander C, Kristiansson E, Hedensti-
doi: 10.1210/jc.2013-4115
erna G, Lindholm C. Expression of the glucocorticoid receptor is
decreased in experimental Staphylococcus aureus sepsis. J Infect.
2013;67:574 –583.
81. Guerrero J, Gatica HA, Rodriguez M, Estay R, Goecke IA. Septic
serum induces glucocorticoid resistance and modifies the expression
of glucocorticoid isoforms receptors: a prospective cohort study and
in vitro experimental assay. Crit Care. 2013;17:R107.
82. Indyk JA, Candido-Vitto C, Wolf IM, et al. Reduced glucocorticoid
receptor protein expression in children with critical illness. Horm
Res Paediatr. 2013;79:169 –178.
83. Peeters RP, Hagendorf A, Vanhorebeek I, et al. Tissue mRNA ex-
pression of the glucocorticoid receptor and its splice variants in fatal
critical illness. Clin Endocrinol (Oxf). 2009;71:145–153.
84. Siebig S, Meinel A, Rogler G, et al. Decreased cytosolic glucocorti-
coid receptor levels in critically ill patients. Anaesth Intensive Care.
2010;38:133–140.
85. Cohen J, Ward G, Prins J, Jones M, Venkatesh B. Variability of
cortisol assays can confound the diagnosis of adrenal insufficiency
in the critically ill population. Intensive Care Med. 2006;32:1901–
1905.
86. Vassiliadi DA, Ilias I, Tzanela M, et al. Interstitial cortisol obtained
by microdialysis in mechanically ventilated septic patients: correla-
tions with total and free serum cortisol. J Crit Care. 2013;28:158 –
165.
87. Venkatesh B, Morgan TJ, Cohen J. Interstitium: the next diagnostic
and therapeutic platform in critical illness. Crit Care Med. 2010;
38:S630 –S636.
88. Chung TT, Grossman A, Clark AJ. Adrenal insufficiency. In: Jame-
son JL, De Groot LJ, eds. Endocrinology: Adult and Pediatric. St
Louis, MO: WB Saunders; 2010;1853–1863.
89. Annane D, Sébille V, Charpentier C, et al. Effect of treatment with
low doses of hydrocortisone and fludrocortisone on mortality in
patients with septic shock. JAMA. 2002;288:862– 871.
90. Bollaert PE, Charpentier C, Levy B, Debouverie M, Audibert G,
Larcan A. Reversal of late septic shock with supraphysiologic doses
of hydrocortisone. Crit Care Med. 1998;26:645– 650.
91. Briegel J, Forst H, Haller M, et al. Stress doses of hydrocortisone
reverse hyperdynamic septic shock: a prospective, randomized, dou-
ble-blind, single-center study. Crit Care Med. 1999;27:723–732.
92. Sprung CL, Annane D, Keh D, et al. Hydrocortisone therapy for
patients with septic shock. N Engl J Med. 2008;358:111–124.
93. Hermans G, Wilmer A, Meersseman W, et al. Impact of intensive
insulin therapy on neuromuscular complications and ventilator de-
pendency in the medical intensive care unit. Am J Respir Crit Care
Med. 2007;175:480 – 489.
94. Hermans G, De Jonghe B, Bruyninckx F, Van den Berghe G. Clinical
review: critical illness polyneuropathy and myopathy. Crit Care.
2008;12:238.
95. Hauer D, Weis F, Papassotiropoulos A, et al. Relationship of a com-
mon polymorphism of the glucocorticoid receptor gene to traumatic
memories and posttraumatic stress disorder in patients after inten-
sive care therapy. Crit Care Med. 2011;39:643– 650.
96. Ledderose C, Möhnle P, Limbeck E, et al. Corticosteroid resistance
in sepsis is influenced by microRNA-124 –induced downregulation
of glucocorticoid receptor-␣. Crit Care Med. 2012;40:2745–2753.
97. Baker AC, Chew VW, Green TL, et al. Single nucleotide polymor-
phisms and type of steroid impact the functional response of the
human glucocorticoid receptor. J Surg Res. 2013;180:27–34.
98. Ahrens CL, Barletta JF, Kanji S, et al. Effect of low-calorie parenteral
nutrition on the incidence and severity of hyperglycemia in surgical
patients: a randomized, controlled trial. Crit Care Med. 2005;33:
2507–2512.
99. Olveira G, Tapia MJ, Ocón J, et al. Parenteral nutrition-associated
hyperglycemia in non-critically ill inpatients increases the risk of
in-hospital mortality (multicenter study). Diabetes Care. 2013;36:
1061–1066.
jcem.endojournals.org 1581
100. Pasquel FJ, Smiley D, Spiegelman R, Lin E, Peng L, Umpierrez GE.
Hyperglycemia is associated with increased hospital complications
and mortality during parenteral nutrition. Hosp Pract. 2011;39:
81– 88.
101. Pieracci F, Hydo L, Eachempati S, Pomp A, Shou J, Barie PS.
Higher body mass index predicts need for insulin but not hyper-
glycemia, nosocomial infection, or death in critically ill surgical
patients. Surg Infect (Larchmt). 2008;9:121–130.
Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
102. Weiss M, Kron M, Hay B, Taenzer M, Radermacher P, Georgieff
M. Which variables are associated with blood glucose levels out-
side the target range in surgical critically ill patients? A retrospec-
tive observational study. Patient Saf Surg. 2011;5:5.
103. Krinsley JS. Glycemic control in the critically ill - 3 domains and
diabetic status means one size does not fit all! Crit Care. 2013;17:
131.
104. Falciglia M, Freyberg RW, Almenoff PL, D’Alessio DA, Render
ML. Hyperglycemia-related mortality in critically ill patients varies
with admission diagnosis. Crit Care Med. 2009;37:3001–3009.
105. van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin
therapy in critically ill patients. N Engl J Med. 2001;345:1359 –
1367.
106. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin
therapy in the medical ICU. N Engl J Med. 2006;354:449 – 461.
107. Vlasselaers D, Milants I, Desmet L, et al. Intensive insulin therapy
for patients in paediatric intensive care: a prospective, randomised
controlled study. Lancet. 2009;373:547–556.
108. Mesotten D, Gielen M, Sterken C, et al. Neurocognitive develop-
ment of children 4 years after critical illness and treatment with
tight glucose control: a randomized controlled trial. JAMA. 2012;
308:1641–1650.
109. Agus MS, Steil GM, Wypij D, et al. Tight glycemic control versus
standard care after pediatric cardiac surgery. N Engl J Med. 2012;
367:1208 –1219.
110. Gunst J, Derese I, Aertgeerts A, et al. Insufficient autophagy con-
tributes to mitochondrial dysfunction, organ failure, and adverse
outcome in an animal model of critical illness. Crit Care Med.
2013;41:182–194.
111. Langouche L, Vanhorebeek I, Vlasselaers D, et al. Intensive insulin
therapy protects the endothelium of critically ill patients. J Clin
Invest. 2005;115:2277–2286.
112. Vanhorebeek I, De Vos R, Mesotten D, Wouters PJ, De Wolf-
Peeters C, Van den Berghe G. Protection of hepatocyte mitochon-
drial ultrastructure and function by strict blood glucose control
with insulin in critically ill patients. Lancet. 2005;365:53–59.
113. Vanhorebeek I, Gunst J, Ellger B, et al. Hyperglycemic kidney
damage in an animal model of prolonged critical illness. Kidney
Int. 2009;76:512–520.
114. Garber AJ, Moghissi ES, Bransome ED Jr, et al. American College
of Endocrinology position statement on inpatient diabetes and
metabolic control. Endocr Pract. 2004;10:77– 82.
115. Umpierrez GE, Hellman R, Korytkowski MT, et al. Management
of hyperglycemia in hospitalized patients in non-critical care set-
ting: an Endocrine Society Clinical Practice Guideline. J Clin En-
docrinol Metab. 2012;97:16 –38.
116. Finfer S, Chittock DR, Su SY, et al. Intensive versus conventional
glucose control in critically ill patients. N Engl J Med. 2009;360:
1283–1297.
117. Finfer S, Liu B, Chittock DR, et al. Hypoglycemia and risk of death
in critically ill patients. N Engl J Med. 2012;367:1108 –1118.
118. Finfer S, Billot L. Managing blood glucose in critically ill patients
with or without diabetes. Crit Care. 2013;17:134.
119. Kosiborod M, Inzucchi SE, Goyal A, et al. Relationship between
spontaneous and iatrogenic hypoglycemia and mortality in pa-
tients hospitalized with acute myocardial infarction. JAMA. 2009;
301:1556 –1564.
120. Tin W, Brunskill G, Kelly T, Fritz S. 15-year follow-up of recurrent
“hypoglycemia” in preterm infants. Pediatrics. 2012;130:e1497–
e1503.
1582 Boonen and Van den Berghe Novel Insights on Endocrine Changes in the ICU J Clin Endocrinol Metab, May 2014, 99(5):1569 –1582

121. Suh SW, Gum ET, Hamby AM, Chan PH, Swanson RA. Hypo-
glycemic neuronal death is triggered by glucose reperfusion and
activation of neuronal NADPH oxidase. J Clin Invest. 2007;117:
910 –918.
122. Van den Berghe G, Wouters PJ, Bouillon R, et al. Outcome benefit
of intensive insulin therapy in the critically ill: insulin dose versus
glycemic control. Crit Care Med. 2003;31:359 –366.
123. Van den Berghe G, Wilmer A, Milants I, et al. Intensive insulin
124. Kavanagh BP, McCowen KC. Clinical practice. Glycemic control
in the ICU. N Engl J Med. 2010;363:2540 –2546.
125. Vanhorebeek I, Langouche L, Van den Berghe G. Endocrine as-
pects of acute and prolonged critical illness. Nat Clin Pract Endo-
crinol Metab. 2006;2:20 –31.
126. Mebis L, Van den Berghe G. Thyroid axis function and dysfunction
in critical illness. Best Pract Res Clin Endocrinol Metab. 2011;25:
745–757.

Downloaded from https://academic.oup.com/jcem/article/99/5/1569/2537306 by National Science & Technology Library user on 10 March 2023
therapy in mixed medical/surgical intensive care units: benefit ver- 127. Van den Berghe G. Intensive insulin therapy in the ICU–reconciling
sus harm. Diabetes. 2006;55:3151–3159. the evidence. Nat Rev Endocrinol. 2012;8:374 –378.

Members have FREE online access to current endocrine

Clinical Practice Guidelines.
www.endocrine.org/guidelines