BRIEF REPORT

Influence of DHEA Administration on

24-Hour Cortisol Concentrations
Patricia D. Kroboth, PHD*†, Janet A. Amico, MD,*†‡ Roslyn A. Stone, PHD*§, Maggie Folan, BSN*†,
Reginald F. Frye, PHD*†, Frank J. Kroboth, MD*‡, Kristin L. Bigos, BS*†, Tanya J. Fabian, PHARMD*†¶,
Ana M. Linares, MS*, Bruce G. Pollock, MD, PHD¶, and Charles Hakala, RPH

dian rhythm similar to cortisol with peak and nadir con-

Abstract: DHEA is marketed and readily available as a daily
centrations occurring during the early light and dark hours,
nutritional supplement to counteract the effects of aging. The ef-
respectively.1–4 DHEA and its sulfate ester, DHEA-S, are
fect of DHEA administration on 24-hour plasma cortisol profiles
both released in response to adrenocorticotropin, ACTH,1
has not been investigated. In this single-blind placebo-controlled
with DHEA-S being the most abundant product of human
crossover study, the effect of DHEA administration on cortisol
steroid biosynthesis. After oral administration, DHEA has
concentrations was evaluated in healthy older women and men.
a half-life of approximately 6 to 12 hours, with women hav-
Once each morning, subjects took either placebo (Days 1 to 7, and
ing a longer half-life than men.{3495} DHEA can be me-
23 to 29) or oral DHEA 200 mg (Days 8 to 22: doses 1 to 15).
tabolized to its sulfate DHEA-S, or converted to andro-
Twenty-four hour DHEA and cortisol concentrations were meas-
stenedione, testosterone, or estradiol.{2431} Cushing’s
ured on Day 1 (placebo), Day 8 (DHEA dose 1), Day 15 (DHEA
syndrome2,5 and some chronic illnesses6 are characterized
dose 8), Day 22 (DHEA dose 15), and Day 29 (placebo washout
by hypersecretion of cortisol relative to DHEA and DHEA-
dose 7). DHEA administration resulted in a decrease in plasma
S, whereas in depression7 and cigarette smoking,8 both cor-
cortisol concentrations (mean, peak, and/or AUC) in healthy
tisol and DHEA and/or DHEA-S concentrations are in-
older women and men. The cortisol-lowering effect of DHEA was
creased. In the present study we evaluated the effects of
more pronounced in women than in men in our study; pairwise
DHEA administration on 24-hour cortisol concentrations
differences in concentrations between days showed that relative
in healthy older individuals.
to Day 1, cortisol was lower on Days 15, 22, and 29 in women (p

0.0001) and on Day 15 in men (p
0.002). The mechanism by METHODS
which DHEA lowers plasma cortisol concentrations merits fur-
Seven women (69.4  3.3 years) and six men (69.7 
ther investigation.
3.3 years) participated in this single-blind placebo-controlled
( J Clin Psychopharmacol 2003;23: 96–99) crossover study of the pharmacokinetics of DHEA.9 This
study was approved by the Institutional Review Board of the

A lthough dehydroepiandrosterone (DHEA) is marketed

as a nutritional supplement to counteract the effects of
aging, its effect upon 24-hour plasma cortisol profiles in
healthy older women and men has not been investigated.
DHEA is synthesized in the human adrenal cortex, is se-
creted synchronously with cortisol,1–3 and exhibits a circa-
*Pharmacodynamic Research Center, †Department of Pharmaceuti-
cal Sciences, ‡Department of Medicine, §Department of Biostatistics, and
¶Department of Psychiatry, University of Pittsburgh, Pittsburgh, Penn-
sylvania; and
Belmar Pharmacy, Lakewood, Colorado
Received February 4, 2002; accepted June 13, 2002.
Research supported by MH-55756, MH-52247 and NIH/NCRR/
GCRC #5M01RR00056, and T32–19986.
Address reprint requests to: Patricia D. Kroboth, PhD, 904 Salk Hall,
Department of Pharmaceutical Sciences, University of Pittsburgh, 3501
Terrace Street, Pittsburgh, PA 15261. Address e-mail to: krobothpd@msx.
upmc.edu
10.1097/01.jcp.0000057196.35767.05
University of Pittsburgh, and all subjects gave written in-
formed consent to participate. All subjects were determined
to be healthy by medical history and examination, bio-
chemical screening and the Structured Clinical Interview for
Diagnosis of DSM IV Disorders. Women were not taking es-
trogen or progesterone supplements.
Once each morning, subjects took either placebo
(Days 1 to 7 and 23 to 29) or DHEA 200 mg (Days 8 to 22:
doses 1 through 15). Identically appearing tablets of mi-
cronized DHEA and placebo were formulated in a wax-
vegetable-oil matrix and compounded in a silica-based ex-
cipient (Belmar Pharmacy, Lakewood, CO). Subjects were
admitted to the General Clinical Research Center (GCRC)
the evening before each of the evaluation days (Days 1, 8,
15, 22, and 29). Subjects ate a light breakfast at 7:30 AM,
and at 8 AM (time 0) took the dose of placebo or DHEA
with water. Plasma DHEA and cortisol concentrations

96 Journal of Clinical Psychopharmacology • Volume 23, Number 1, February 2003

Journal of Clinical Psychopharmacology • Volume 23, Number 1, February 2003 DHEA Administration and Cortisol Concentrations

were measured after the “time 0” sample, and were ob-

tained on Day 1 (placebo), Day 8 (DHEA dose 1), Day 15
(DHEA dose 8), Day 22 (DHEA dose 15), and Day 29
(placebo washout dose 7). On Day 22, there were data
from only five men. Blood samples (6 mL each) were ob-
tained from a forearm vein through an indwelling catheter
at 0 h and at 1, 2.5, 3.5, 4, 5, 6, 7, 9, 11, 14, and 23.5 h af-
ter study drug administration; samples were centrifuged
and plasma was separated and stored at 80 C until as-
sayed. After collection of the 23.5 h blood sample, subjects
were discharged and, except for Day 29, given the appro-
priate tablets (either placebo or DHEA) for the subsequent
six days. Plasma cortisol and DHEA concentrations were
measured by 125I-radioimmunoassay kits (Diagnostic Sys-
tems Laboratories, Webster TX).
The repeated cortisol measurements from women
and men on each of the four days were analyzed separately
using a mixed-model analysis of variance (ANOVA), with
day, time, and the day*time interaction as fixed effects and
subject as a random effect. To determine pairwise differ-
ences in cortisol concentrations between days, least squares
means was used as the post hoc analysis. SAS® software was
used.10 Preliminary analysis of cortisol data revealed that
the day*time interaction was not statistically significant for
either women or men (p  0.5 for each) and was dropped
from the model. Area under the curve (AUC) for 0 to 23.5
h was determined by the trapezoidal rule; peak cortisol
concentration data (0 h) were also analyzed using a mixed-
model approach, but without the time term. Post-hoc pair-
wise comparisons between the days were considered statis-
tically significant at a level of p  0.05.

RESULTS
As shown in the two panels of Figure 1, cortisol dis-
played a diurnal pattern of secretion during placebo (Day
1) and during administration of DHEA (Days 15 and 22
shown). Though data are not shown, the pattern was simi-
lar on the first day of DHEA administration (Day 8) and af-
ter placebo washout (Day 29). Figure 1 also demonstrates
that DHEA administration significantly decreased cortisol
FIG. 1. Mean cortisol concentrations at 0 h (8 AM) and at 1,
concentrations in both women and men.
2.5, 3.5, 4, 5, 6, 7, 9, 11, 14, and 23.5 h after study drug ad-
In these seven older women, there was a significant ministration in (1A) seven women on Day 1 (placebo) and Day
difference in cortisol concentrations among the four ob- 22 (DHEA dose 15) and (1B) six men on Day 1 and Day 15
servation days (p
0.0001). Figure 1A shows concentra- (DHEA dose 8).
tions in women on Day 1 and Day 22. Least squares means
post-hoc analysis to determine pairwise differences in con- evaluation days. Cortisol concentrations on Day 15 were
centrations between days showed that relative to Day 1, also significantly lower than on each of the other days (p 
cortisol was significantly lower on Days 15, 22, and 29 0.05, for each).
(p
0.0001, for each) and tended to be lower on Day 8 Peak morning plasma cortisol (0 h) in women dif-
(p
0.08). fered among the days (p  0.04) and was significantly
In these six older men, cortisol concentrations were lower on Day 15 (DHEA dose 8) than on Day 1 (p
0.02).
significantly lower on Day 15 (dose 8) than on Day 1 (p
No statistically significant differences across the days were
0.002, Figure 1B). A similar trend was noted on the other observed in peak cortisol for men (p
0.25), or for nadir

© 2003 Lippincott Williams & Wilkins 97

Kroboth and Associates Journal of Clinical Psychopharmacology • Volume 23, Number 1, February 2003
plasma cortisol concentrations for either women (p
0.88)
or men (p
0. 42).
Figure 2 shows mean AUCs for cortisol for women
(2A) and men (2B). Cortisol AUCs differed among the
days for women (p  0.04), but not for men (p  0.11).
Among the women, cortisol AUCs were statistically lower
on Days 15, 22, and 29 than on Day 1 (p  0.03), and
tended to be lower on Day 8 (p
0.07). Figure 2 also
shows the mean ratio of DHEA AUC to cortisol AUC on
each evaluation day, for women (2A) and men (2B).
DHEA administration resulted in mean increases of 703%,
638%, and 678% in the DHEA/cortisol ratio on Days 8, 15,
FIG. 2. Mean cortisol AUCs are represented as the bars plot-
ted on the left axis for each evaluation day in seven women
(2A) and six men (2B). The mean ratio of DHEA AUC to corti-
sol AUC is represented as the line plotted on the right axis for
each evaluation day in women (2A) and men (2B). Statistical
significance (p  0.03 for all) is denoted by *. On Day 22, there
are data for only five men.
98
and 22, respectively, in women; in men, the increases were
419%, 499%, and 475%.
As previously described,9 endogenous plasma con-
centrations of DHEA (Day 1) were similar in women and
men (p
0.99). DHEA concentrations were higher after
DHEA administration (Days 8, 15, and 22) than after
placebo (Days 1 and 29) in both women and men (p 
0.0001). DHEA concentrations were higher in women
than in men (p  0. 03) during DHEA administration
(Days 8, 15, 22). On placebo washout dose 7 (Day 29),
DHEA concentrations were similar to Day 1 concentra-
tions in women (p
0.91) and men (p
0.93).
DISCUSSION
This study demonstrates that DHEA administration
results in a decrease in plasma cortisol concentrations
(peak, mean, and/or AUC) in healthy older individuals.
This observation suggests that the anti-glucocorticoid ac-
tivity of DHEA is two-fold: first, through its intrinsic phar-
macological effects,14 and second, through the cortisol low-
ering effect observed in this study.
The overall diurnal pattern in plasma cortisol was
preserved before, during, and after a one-week washout of
DHEA administration, but at a decreased level during and
after DHEA administration. Prior reports have noted de-
creased plasma cortisol concentrations after a single dose
of DHEA.11 Labrie and associates12 measured serum cor-
tisol in single blood samples obtained before, immediately
after, and on days 3, 7, 11, and 14 of percutaneous DHEA
administration to 60 to 70 year old men and women, but
found no effect on cortisol. However, both the small num-
ber of sampling times and a different route of administra-
tion of DHEA may account for the different outcome than
in the present study.
One possible explanation for the cortisol-lowering ef-
fect of DHEA is that DHEA and/or DHEA-S may have an
inhibitory effect on the activity of the hypothalamic-
pituitary-adrenal axis. DHEA is not believed to directly in-
hibit either ACTH or cortisol release; however, its effects
upon corticotropin-releasing factor and arginine vasopressin,
the hypothalamic factors that regulate ACTH secretion
from the pituitary, have not been investigated.2 While an in-
crease in cortisol metabolic clearance is a possibility, the fact
that DHEA-S inhibits cytochrome P4503A4 (CYP3A4)13
and therefore cortisol metabolism, indicates that an increase,
not a decrease, in cortisol concentrations would occur dur-
ing DHEA administration. Therefore, a change in cortisol
clearance is an unlikely mechanism.
The cortisol-lowering effect of DHEA was more pro-
nounced in women than in men in our study, and cannot
be explained by differences in age or basal levels of
DHEA. However, higher concentrations of DHEA (after
DHEA administration) were achieved in women than in
© 2003 Lippincott Williams & Wilkins
Journal of Clinical Psychopharmacology • Volume 23, Number 1, February 2003 DHEA Administration and Cortisol Concentrations
men,9 which may explain the significant DHEA-lowering
effect in women, but not in men. The persistently lower
plasma cortisol concentrations on Day 29, which followed
a 7-day washout with placebo, are also consistent with our
observation that although DHEA concentrations had re-
turned to basal levels in both men and women, DHEA-S
concentrations had returned to baseline in men only.9 In
men, significantly lower cortisol concentrations were iden-
tified only on the eighth day of DHEA administration.
Similar changes in cortisol were observed on the other
days, though these decreases did not reach statistical sig-
nificance.
The anti-glucocorticoid properties of DHEA may
have potential therapeutic value in states in which DHEA
declines15 or glucocorticoids are either secreted or admin-
istered in supraphysiological amounts. For example, de-
spite hypercortisolism in depressed subjects7 or cigarette
smokers,8 signs of glucocorticoid excess do not develop,
perhaps because of the coexistent increase in DHEA.
Therefore, the relative abundance of DHEA to cortisol may
be biologically more relevant than the absolute concentra-
tion of either hormone alone. The mechanism by which
DHEA lowers plasma cortisol concentrations merits fur-
ther investigation due to the potential clinical implications
resulting from an increase or decrease in these hormones.16
ACKNOWLEDGMENTS
The authors thank the nurses and staff of the Uni-
versity of Pittsburgh General Clinical Research Center,
University of Pittsburgh, and Ms. Deborah Hollingshead,
for their expert technical assistance.
REFERENCES
1. Kroboth PD, Salek FS, Pittenger AL, et al. DHEA and DHEA-S: A
review. J Clin Pharmacol 1999;39:1–22.
© 2003 Lippincott Williams & Wilkins
2. Orth DN, Kovacs WJ. The adrenal cortex, in: Wilson JD, Foster DW,
Kronenberg HM, Larsen PR (ed.), Williams Textbook of Endo-
crinology. Philadelphia: W. B. Saunders Company, 1998;517–664.
3. Rosenfeld RS, Hellman L, Roffwarg H, et al. Dehydroisoandros-
terone is secreted episodically and synchronously with cortisol by
normal man. J Clin Endocr 1971;33:87–92.
4. Weitzman ED, Fukushima C, Nogeire C. Twenty-four hour pattern
of the episodic secretion of cortisol in normal subjects. J Clin En-
docrinol Metab 1971;33:14–22.
5. Parker LN, Odell WD. Control of adrenal androgen secretion. En-
docr Rev 1980;1:392–10.
6. Parker LN, Levin ER, Lifrak ET. Evidence for adrenocortical adap-
tation to severe illness. J Clin Endocrinol Metab 1985;60:947–52.
7. Heuser I, Deuschle M, Luppa P, et al. Increased diurnal plasma con-
centrations of dehydroepiandrosterone in depressed patients. J Clin
Endocrinol Metab 1998;83:3130–33.
8. Khaw K-T, Chir MBB, Tazuke S, et al. Cigarette smoking and levels
of adrenal androgens in postmenopausal women. New Engl J Med
1988;318:1705–09.
9. Frye RF, Kroboth PD, Kroboth FJ, et al. Sex differences in pharma-
cokinetics of dehydroepiandrosterone (DHEA) after single- and
multiple-dose administration in healthy older adults. J Clin Pharma-
col 2000;40:596–605.
10. SAS Institute I. The MIXED Procedure, in: SAS Technical Report P-
229, SAS/STATR Software: Changes and Enhancements, Release
6.07. Cary, NC: SAS Institute Inc., 1992;289–366.
11. Wolf OT, Koster B, Kirschbaum C, et al. A single administration of
dehydroepiandrosterone does not enhance memory performance in
young healthy adults, but immediately reduces cortisol levels. Biol
Psychiatry 1997;42:845–8.
12. Labrie F, Bélanger A, Cusan L, et al. Physiological changes in dehy-
droepiandrosterone are not reflected by serum levels of active an-
drogens and estrogens but of their metabolites: Intracrinology. J Clin
Endocrinol Metab 1997;82:2403–9.
13. Frye RF, Kroboth PD, Folan MM, et al. Effect of DHEA on CYP3A-
mediated metabolism of triazolam. Clin Pharmacol Ther 2000;67:109.
14. Wright BE, Porter JR, Browne ES, et al. Antiglucocorticoid action of
dehydroepiandrosterone in young obese Zucker rats. Int J Obes
1992;16:579–83.
15. Morales AJ, Nolan JJ, Nelson JC, et al. Effects of replacement dose
of dehydroepiandrosterone in men and women of advancing age. J
Clin Endocrinol Metab 1994;78:1360–7.
16. Labrie F, Bélanger A, Cusan L, et al. Marked decline in serum con-
centrations of adrenal C19 sex steroid precursors and conjugated an-
drogen metabolites during aging. J Clin Endocrinol Metab 1997;82:
2396–2402.
99