A Healthy Body in a Healthy Mindand Vice VersaThe Damaging Power of "Uncontrollable" Stress
George P. Chrousos and
Philip W. Gold
National Institutes of Health
Bethesda, Maryland 20892
Address correspondence and requests for reprints to: George P. Chrousos, M.D., NIH, Building 10, Room 10N262, 10 Center Drive MSC 1862, Bethesda, MD 20892-1862. E-mail: George_Chrousos{at}NIH.Gov
 |
Introduction
|
---|
The hypothalamic-pituitary-adrenal (HPA)
axis together with the sympathetic system connect the brain with the
periphery of the body (1) (Fig. 1
). The
central nervous system (CNS) centers of the HPA axis and the
sympathetic system are, respectively, the parvicellular
corticotropin-releasing hormone (CRH) and arginine-vasopressin (AVP)
neurons of the paraventricular nuclei of the hypothalamus and the
noradrenergic neurons of the locus ceruleus/norepinephrine (LC/NE)
nuclei of the brain stem. These neuronal centers innervate and
stimulate each other and have both a baseline circadian and
stress-related activity. The CRH/AVP and LC/NE neurons and their
peripheral axes are heuristically known as the stress system. The
secretion of the end-product of the HPA axis, cortisol, is kept by an
elaborate negative feedback system within an optimal time-integrated
narrow range, which is quite stable in an individual subject. That the
body would have such a tightly regulated servo-control system suggests
that excessive, unchecked activity could be detrimental to the
organism.

View larger version (24K):
[in this window]
[in a new window]
|
Figure 1. A schematic representation of the stress
system. The CRH/AVP neurons are reciprocally connected with the
noradrenergic neurons of the LC/NE system in a positive reverberatory
circuit. The HPA axis is controlled by several negative feedback loops,
which tend to normalize the time-integrated secretion of cortisol, yet
glucocorticoids stimulate the amygdala and, hence, the fear center.
Activation of the HPA axis leads to suppression of the GH/IGF-1,
LH/testosterone/estradiol, and TSH/T3 axes; activation of
the sympathetic system increases IL-6 secretion. Chronic increases in
cortisol, catecholamines, and IL-6 and chronic suppression of the
GH/IGF-1, LH/T and TSH/T3 axes lead to visceral obesity,
hypertension, atherosclerosis, osteoporosis, and immune dysfunction and
their sequelae resulting in increased morbidity and mortality. Symbols:
Solid lines indicate stimulation; interrupted
lines indicate inhibition. Abbreviations: HPA,
hypothalamic-pituitary-adrenal; CRH, corticotropin-releasing hormone;
AVP, arginine-vasopressin; LC/NE, locus ceruleus/norepinephrine system;
GH, growth hormone; IGF-1, insulin-like growth factor-1; LH,
luteinizing hormone; T, testosterone; TSH, thyrotropin; T3,
triiodothyronine; F, cortisol; NE, norepinephrine; E, epinephrine;
IL-6, interleukin-6.
|
|
Indeed, excessive and sustained cortisol secretion or chronic
pharmacologic doses of glucocorticoids (endogenous or exogenous Cushing
syndrome, respectively) have been long associated with depression,
hypertension, osteoporosis, immunosuppression and the entire spectrum
of metabolic syndrome X, including visceral obesity, insulin
resistance, dyslipidemia, dyscoagulation, and hypertension, along with
their morbid sequelae of atherosclerosis and cardiovascular disease (2, 3) (Fig. 1
, Table 1
). Each and every one
of these manifestations could in theory be produced, despite the
presence of normal, nonhyperfunctioning HPA axis, by tissue-specific
hypersensitivity to glucocorticoids of, respectively, the amygdala or
mesocorticolimbic system, cardiovascular system, bone, immune system,
or adipose tissue (4, 5). The papers by Rosmond et al.
(6) and Panarelli et al. (7) in this issue of
JCEM (see pages 1853 and 1846) represent pioneering attempts
to respectively define whether normal life stress-related
hypersecretion of cortisol or tissue-limited hypersensitivity to
glucocorticoids could affect important physiologic parameters, such as
systemic blood pressure, or functions, such as carbohydrate and
lipid metabolism, with potential deleterious effects on the organism in
the long-term.
After the seminal discovery of CRH in the early 80s and following a
series of studies in experimental animals and patients with melancholic
depression, we described the marked clinical, physiologic, and
biochemical similarities between acute stress and the melancholic
depression syndrome (8, 9). Both conditions are indeed associated with
a hyperactive HPA axis and LC/NE system and, hence, with increased CRH,
cortisol, and catecholamine secretion, plus consequent inhibition of
the growth, thyroid, and reproductive axes, suppression of the immune
system, and elevation of catecholamine-stimulated interleukin-6 (IL-6)
concentrations (10). In the case of melancholic depression, the
hyperactivity of the stress system can be chronic or in repeated bouts,
which could potentially produce the long-term consequences of Cushing
syndrome (Fig. 1
, Table 1
). Indeed, we recently demonstrated that
earlier history of melancholic depression was associated with marked
osteoporosis in premenopausal women carefully matched for body mass
index (BMI) to premenopausal controls (11). Furthermore, patients with
depressive symptomatology, including properly diagnosed melancholic
depression, have a markedly decreased life expectancy due to increased
mortality from primarily cardiovascular causes (relative risk 23 over
gender- and age-matched controls) (12, 13, 14). Although only 1015% of
the adult population may fulfill the criteria for major depression, it
is quite likely that there is a continuum of depressive symptomatology,
with only the upper cut of patients qualifying as melancholics.
Rosmond et al. (6) examined a large unselected population of
53-yr-old men by obtaining a detailed history, by performing physical
examinationsincluding anthropometric measurements, by obtaining a
series of diurnal salivary cortisol determinations in parallel with an
acceptable measure of stress perception, and by performing a low-dose
overnight dexamethasone suppression test. They analyzed the results in
a complex, yet quite logical fashion, which revealed that the increases
in blood pressure and body mass index, earlier seen in Cushing syndrome
as a result of hypercortisolism, could also be seen in a general
population of nonCushingoid middle-aged men in correlation with the
degree of stress perception and stress-related cortisol secretion.
A crucial observation was made upon the initial analysis of data,
modeled and extended in Fig. 2A
. A
nonstressed HPA axis was characterized by increased variance, mostly
due to a wide circadian variation, with distant morning zeniths and
evening nadirs, a discrete but small lunch-induced cortisol peak and an
appropriate suppression of the morning cortisol levels in response to
low-dose dexamethasone; a chronically stressed HPA axis, on the other
hand, was characterized by a decreased variance mostly due to evening
nadir elevations and morning zenith decreases, a large lunch-induced
cortisol response and an inadequate suppression of morning cortisol by
overnight dexamethasone. These findings suggest chronic hypersecretion
of CRH in chronically stressed individuals and a reset of their HPA
axis as previously suggested. How about the total time-integrated
cortisol secretion? Is it not important? We are sure that it is;
however, in the presence of a properly functioning glucocorticoid
negative feedback system, around-the-clock cortisol secretion would be
minimized to the greatest extent and, hence, would be less indicative
of a chronically stressed HPA axis than the other features suggested by
Rosmond et al.

View larger version (27K):
[in this window]
[in a new window]
|
Figure 2. A, (Left) Circadian pattern of cortisol
secretion in nonstressed (NS, ) and chronically stressed (CS, .....),
individuals. Note the blunting of the circadian rhythm in the latter,
along with an augmented cortisol elevation in response to lunch.
(Right) Cortisol response to a low dose of overnight
dexamethasone (D). Note the increased suppressibility of nonstressed
individuals vs. chronically stressed subjects. B,
Dose-response curves of target tissue responses to cortisol: N, normal;
HS, hypersensitive; R, resistant. The interrupted horizontal
line represents a threshold effect beyond which long-term harm is
done (Fig. 1 , Table 1 ). Depending on the shift of the dose response
curve to the left or right one would expect harmful or protective
effects.
|
|
The ability of the glucocorticoid negative feedback system to limit the
production of cortisol during stress can be impaired by chronic
emotional or physical stress and by old age (15, 16, 17).
Glucocorticoid-induced hippocampal neuron damage and deficient
transmission of suprahypothalamic negative feedback has been proposed
as a major mechanism mediating this phenomenon (18) (Fig. 1
, Table 1
).
Indeed, in patients with melancholic depression, the 24-h urinary free
cortisol excretion increases with age, while studies of the HPA axis in
aging populations that include persons with chronic emotional or
physical diseases have shown progressive elevations of evening plasma
cortisol concentrations with age.
But can an altered daily cortisol secretion variance result in the
somatic sequelae of chronic hypercortisolism? Despite the attempt of
the brain to correct for the evening excess cortisol production by
suppressing the morning cortisol surge, it is possible that no complete
such correction is attained, and the body tissues are overexposed to
cortisol. On the other hand, blunting of the circadian rhythm could
result in evening exposure to cortisol, which could be detrimental on
its own, in spite of an adequate correction of time-integrated
cortisol secretion. We recently reported an adult man with Carney
complex treated in childhood with unilateral adrenalectomy (19).
Although his 24-h urinary free cortisol excretion remained normal for
many years, he developed severe osteoporosis, possibly as a result of
constant exposure of his bones to "normal" levels of plasma
cortisol. The most impressive data of the Rosmond study (6) are those
described in their Table 5, in which they correlate stress-related
cortisol secretion (
i) corrected for the inverse of the daily
variance (vi) (
=1÷vi) to amplify the effect of the low variance
observed in chronically stressed subjects. One can see all the
correlations one would have expected from the scheme in Fig. 1
and the
effects outlined in Table 1
.
Panarelli et al. (7) studied a smaller group of younger
males, ages 1840 y. They focused their studies on a previously
described polymorphism of the glucocorticoid receptor, which was
earlier associated with hypertension and visceral obesity (20, 21).
This polymorphism was associated with an increased blanching skin
reaction to butesonide, but not with systemic blood pressure, plasma
biochemistries known to be affected by glucocorticoids (Table 1
), the
affinity or concentration of glucocorticoid receptors in cultured
leukocytes, or the dexamethasone-induced inhibition of lysozyme
production by cultured leukocytes in vitro. Thus, these
authors found one hypersensitive dose-response curve to
glucocorticoids-skin vasoconstrictionbut not others (Fig. 2B
).
Whyor why not? The finding of the correlation between an undefined
noncoding polymorphism of the glucocorticoid receptor gene and a
hypersensitive curve is proof that tissue-limited hypersensitivity to
glucocorticoids, or its mirror image glucocorticoid resistance, do
exist, as theoretically hypothesized (4). Huizenga et al.
(22) recently demonstrated in this journal that another polymorphism of
the glucocorticoid receptor, which we described, tested, and found not
to have an effect on function of the receptor in vitro (23),
was present in 6% of normal Dutch men and was associated with a
significantly greater cortisol suppression by dexamethasone, a higher
BMI and a lower bone mineral density (BMD) in polymorphism carriers
than in noncarriers. These findings are compatible with
hypersensitivity of the hippocampus, adipose tissue, and bone to
glucocorticoids in the carriers. Finally, experimental expression of
increased levels of glucocorticoid receptors in the pancreatic
ß-cells of transgenic mice caused defective insulin secretion and
carbohydrate intolerance (24). A similarly defective insulin secretion
has been observed in patients with Cushing syndrome as well (3). Our
opinion is that, within the human population, there is variation in
target-gene-specific responsiveness to glucocorticoids, which is the
result of not only mutations in the gene of the glucocorticoid
receptor, but also in genes that are involved in the glucocorticoid
signal transduction pathway, including cortisol metabolizing enzymes,
heat shock proteins, coactivators/corepressors, etc (5). These normal
variations could be harmful or protective, depending on the gene and
the direction of the variation.
In summary, the studies of Rosmond et al. (6) and Panarelli
et al. (7) have addressed two prongs of a highly complex,
multifactorial, polygenically determined, developmental, and
environmentally-dependent phenomenon of major importance to medicine
and society. The complex picture of this phenomenon has been unraveling
since the fields of stress and depression coalesced, to give us a clear
biological view of this huge area, with the potential to intervene
rationally to both prevent and treat (8, 9). Now the appropriate
changes of lifestyle will be based on solid biomedical evidence, and
the treatment of emotional disorders will also be therapy for
devastating organic diseases. A healthy mind will define a healthy
body, and vice versa.
Received March 17, 1998.
Accepted March 23, 1998.
 |
References
|
---|
-
Chrousos GP, Gold PW. 1992 The concepts of
stress and stress system disorders: Overview of physical and behavioral
homeostasis. JAMA. 267:12441252.[Abstract]
-
Tsigos C, Chrousos GP. 1996 Differential diagnosis
and management of Cushings syndrome. Ann Rev Med. 47:443461.[CrossRef][Medline]
-
Friedman TC, Mastorakos G, Newman TD, et al. 1996 Carbohydrate and lipid metabolism in endogenous hypercortisolism:
shared features with metabolic syndrome X and noninsulin-dependent
diabetes mellitus. Endocr J. 43:645656.[Medline]
-
Chrousos GP, Detera-Wadleigh S, Karl M. 1993 Syndromes of glucocorticoid resistance. Ann Intern Med. 119:11131124.[Abstract/Free Full Text]
-
Bamberger CM, Schulte HM, Chrousos GP. 1996 Molecular determinants of glucocorticoid receptor function and tissue
sensitivity. Endo Rev. 17:221244.[Medline]
-
Rosmond R, Dallman MF, Björntorp P. 1998 Stress-related cortisol secretion in men: Relationships with abdominal
obesity, endocrine, metabolic and hemodynamic abnormalities. J
Clin Endocrinol Metab. 83:18531859.[Abstract/Free Full Text]
-
Panarelli M, Holloway CD, Fraser F, McConnell J, Ingram
MC, Anderson NH, Kenyon CJ. 1998 Glucocorticoid receptor
polymorphism, skin vasoconstriction and other metabolic intermediate
phenotypes in normal human subjects. J Clin Endocrinol Metab. 83:18461852.[Abstract/Free Full Text]
-
Gold PW, Goodwin FK, Chrousos GP. 1988 Clinical
and biomedical manifestations of depression: Relationship to the
neurobiology of stress (Part I). N Engl J Med. 319:348353.[Medline]
-
Gold PW, Goodwin FK, Chrousos GP. 1988 Clinical
and biomedical manifestations of depression: Relationship to the
neurobiology of stress (Part II). N Engl J Med. 319:413420.[Abstract]
-
Papanicolaou DA, Wilder RL, Manolagas SC, Chrousos
GP. 1998 The pathophysiologic roles of interleukin-6 in human
disease. Ann Intern Med. 128:127137.[Abstract/Free Full Text]
-
Michelson D, Stratakis C, Hill LY, Reynolds J, Galliven
E, Chrousos GP, Gold PW. 1996 Bone mineral density is
decreased in women with depression. N Engl J Med. 335:11761181.[Abstract/Free Full Text]
-
Barefoot JC, Scholl MD. 1993 Symptoms of
depression, acute myocardial infarction, and total mortality in a
community sample. ACP J Club 93:19761980.
-
Anda R, Williamson D, Jones C, et al. 1993 Depressed affect, hopelessness, and the risk of ischemic heart disease
in a cohort of U.S. adults. Epidemiology. 4:285294.[Medline]
-
Silverstone PH. 1990 Depression increases mortality
and morbidity in acute life-threatening medical illness. J
Psychosom Res. 34:651657.[CrossRef][Medline]
-
Halbreich U, Asnis GM, Zumoff B, Nathan RS,
Shindledecker R. 1984 Effects of age and sex on cortisol secretion
in depressives and normals. Psychiatry Res. 13:221229.[CrossRef][Medline]
-
Kling MA, Roy A, Doran AR, et al. 1991 Cerebrospinal fluid immunoreactive corticotropin-releasing hormone and
adrenocorticotropin secretion in Cushings disease and major
depression: Potential clinical implications. J Clin Endocrinol
Metab. 72:260271.[Abstract]
-
Van Cauter E, Leproult R, Kupfer DJ. 1996 Effects
of gender and age on the levels of circadian rhythmicity of plasma
cortisol. J Clin Endocrinol Metab. 81:24682473.[Abstract]
-
McEwen BS. Protective and damaging effects of
stress mediators. 1998 N Engl J Med. 338:171179.[Free Full Text]
-
Sarlis NJ, Chrousos GP, Doppman JL, Carney JA, Stratakis
CA. 1997 Primary pigmented nodular adrenocortical disease:
Reevaluation of a patient with Carney Complex 27 years after unilateral
adrenalectomy. J Clin Endocrinol Metab. 82:12741278.
-
Watt GCM, Harrap SB, Roy CJW, Holton DW, Edwards HW,
Davidson HR, Connor JM, Lever AF, Fraser R. 1992 Abnormalities of
glucocorticoid metabolism and the renin-angiotensin system: A four
corners approach to the identification of the genetic determinants of
blood pressure. J Hypertension 10:473482.
-
Weaver JU, Hitman GA, Kopelman PG. 1992 An
association between B11 restriction fragment length polymorphism of the
glucocorticoid receptor locus and hyperinsulinemia in obese women. J
Mol Endocrinol. 9:295300.[Abstract]
-
Huizenga NATM, Koper JW, De Lange P, Pols HAP, Stolk RP,
Burger H, Grobbee DE, Brinkman AO, De Jong FH, Lamberts SWJ. 1998 A polymorphism in the glucocorticoid receptor gene may be associated
with an increased sensitivity to glucocorticoids in vivo. J Clin
Endocrinol Metab. 83:144151.[Abstract/Free Full Text]
-
Karl M, Lamberts SW, Detera-Wadleigh S, Encio IJ,
Stratakis CA, Hurley DM, Accili D, Chrousos GP. 1993 Familial
glucocorticoid resistance caused by a splice site deletion in the human
glucocorticoid receptor gene. J Clin Endocrinol Metab.
76;683689.
-
Delaunay F, Khan A, Cintra A, et al. 1997 Pancreatic ß cells are important targets for the diabetogenic effects
of glucocorticoids. J Clin Invest. 100:20942098.[Abstract/Free Full Text]