Hydrocortisone effects on cardiovascular variability in septic shock:
A spectral analysis approach*
Jerome Aboab, MD; Andrea Polito, MD; David Orlikowski, MD; Tarek Sharshar, MD, PhD;
Muriel Castel; Djillali Annane, MD, PhD
S
eptic shock still places a bur-
den on the healthcare system
(1). The mechanisms of cardio-
vascular dysfunction in sepsis
may involve excessive release of vasodila-
tors such as cytokines, platelet activating
factor, prostacyclin, and nitric oxide. Si-
multaneously, the response of vasocon-
strictor systems such as the renin-
angiotensin or the sympathetic nervous
systems may be inhibited (2). Assessment
of the variability of cardiovascular signals
provides information on the control of
vascular tone and heart by the autonomic
nervous system (3). Then, studies on
spectral analysis of heart rate and blood
pressure signals suggested that a de-
creased sympathetic modulation of the
cardiovascular system could precede
shock in endotoxin-challenged animals
(4) and in patients with severe sepsis (5).
These data together with recent post
mortem findings of neuronal apoptosis in
the cardiovascular autonomic centers ar-
gued for a role of the dysfunction of the
autonomic nervous system in the onset of
shock in sepsis (6). Similarly, impaired
hypothalamic pituitary adrenal axis is a
common complication of severe sepsis
(7–9), and exogenous administration of
hydrocortisone restores the cardiovascu-
lar homeostasis (10, 11). However, the
exact mechanisms underlying the hemo-
dynamic effects of hydrocortisone remain
uncertain. Glucocorticoids inhibit the in-
ducible form of the nitrite oxide synthase
and the cyclooxygenase II via genomic
interactions (12), and likely enhance the
autonomic outflow and the coupling of
the alpha receptor to its agonists (13).
We hypothesized that, in sepsis, the
early vascular effects of hydrocortisone
could result, at least partly, from an im-
provement in the autonomic nervous
control of cardiovascular variability.
METHODS
The study protocol was approved by the
Comité Consultatif de Protection des Per-
sonnes se prêtant à une Recherche Biomédi-
cale de Saint Germain en Laye (France), and
informed consent was obtained from patients’
relatives and healthy controls.
Study Populations. Intensive care unit pa-
tients who met the American College of Chest
Physicans/Society of Critical Care Medicine
Consensus Conference criteria for septic
shock (14) were prospectively included. Exclu-
sion criteria were age 18 yrs, nonsinus
rhythm, pregnancy, acute myocardial infarc-
tion, pulmonary embolism, previous treat-
ment with corticosteroids, known autoim-
mune disease or immune suppression,
chronic cardiovascular, pulmonary or neuro-
logic diseases (e.g., polyneuropathy, traumatic
brain injuries, cerebral hemorrhages, or isch-
emia, and chronic neurodegenerative dis-
eases), diabetes mellitus, and any other condi-
tion that may be associated with autonomic
failure.
Six nonsmoking volunteers were included
if they had normal thorough medical exami-
*See also p. 1658.
From Réanimation Polyvalente, Hôpital Raymond
Poincaré, AP-HP, Université de Versailles Saint Quen-
tin, Garches, France.
Supported by a grant from the Sepsis Academy of
Ile de France (SAIF).
The authors have not disclosed any potential con-
flicts of interest.
For information regarding this article, E-mail:
djillali.annane@rpc.aphp.fr
Copyright © 2008 by the Society of Critical Care
Medicine and Lippincott Williams & Wilkins
DOI: 10.1097/CCM.0b013e31816f48f2
Rational: Septic shock may be associated with a loss in
cardiovascular variability and adrenal dysfunction.
Objectives: To investigate the relationship between cardiovas-
cular autonomic modulation and adrenal function during sepsis.
Measurement and Main Results: Seventy-five volunteers with
septic shock and six healthy volunteers were prospectively in-
cluded in the study. Cardiovascular variability was assessed by
spectral analysis of heart rate and diastolic blood pressure sig-
nals, which included computation of normalized low (LF
nu
) and
high frequency (HF
nu
) components. Cardiovascular variability was
investigated in patients and healthy volunteers immediately be-
fore and 1 hr after a single bolus of 50 mg of hydrocortisone
(study phase I); in patients according to adrenal function (study
phase II); and in patients with septic shock and adrenal insuffi-
ciency, before and 72 hrs after a treatment with 50 mg every 6 hrs
of hydrocortisone and 50 g daily of fludrocortisone or their
placebos (study phase III). As compared to healthy volunteers,
patients had decreased LF
nu
-HR (.16 .05 vs. .23 .07 p .01)
and LF
nu
-DBP (.18 .11 vs. .28 .02 p .01) and, after
hydrocortisone, they had a greater increase in LF
nu
-DBP (p .01).
As compared to patients with normal adrenal function, those
with adrenal failure had decreased LF
nu
-HR (.1 .01 vs. .2
.15 p .01) and LF
nu
-DBP (.008 .01 vs. .14 .22 p .0003).
In patients with adrenal failure, as compared to placebos,
hydrocortisone plus fludrocortisone increased significantly
LF
nu
-DBP (p .02) and low frequency/high volume ratio (p
.009).
Conclusion: In septic shock, the loss in cardiovascular vari-
ability is more marked in patients with adrenal insufficiency
and is partly restored by exogenous administration of cortico-
steroids. (Crit Care Med 2008; 36:1481–1486)
KEY WORDS: septic shock; cardiovascular variability; hydrocor-
tisone; fludrocortisone; autonomic dysfunction; sympathetic ner-
vous system; parasympathetic nervous system
1481 Crit Care Med 2008 Vol. 36, No. 5