0021-972X/98/$03.00/0
Journal of Clinical Endocrinology and Metabolism
Copyright © 1998 by The Endocrine Society
Vol. 83, No. 2
Printed in U.S.A.
COMMENTS
High Fluid Intake Increases Urine Free Cortisol
Excretion in Normal Subjects
M. VERONICA MERICQ
AND
GORDON B. CUTLER, JR.*
Developmental Endocrinology Branch, National Institutes of Child Health and Human Development,
National Institutes of Health, Bethesda, Maryland 10892-1862
ABSTRACT
To test the hypothesis that increased fluid intake increases the
urine free cortisol, we prescribed 5 liters of fluid intake per day or a
normal fluid intake according to a randomized cross-over design in six
normal volunteers. Each period lasted 5 days, with a 2-day washout
period of normal fluid intake between the two periods. Urine free
cortisol, 17-hydroxycorticosteroids, and creatinine were measured
daily during each study period, and the average value over each 5-day
period was calculated for each subject. High fluid intake caused a
significant increase in the mean urine free cortisol [126 6 33 (SD) vs.
77 6 18 mg/day, P , 0.005]. The frequency of urine free cortisol results
that exceeded the upper normal limit of 95 mg/day was also much
higher during high fluid intake (23/30 vs. 6/30, P , 0.005). By contrast, urine 17-hydroxycorticosteroids (high fluid vs. normal fluid:
5.3 6 1.5 vs. 5.0 6 1.7 mg/day, respectively, P 5 not significant) and
urine creatinine (1.51 6 0.48 vs. 1.45 6 0.37 g/day, P 5 not significant)
did not differ between the two study periods. We conclude that high
fluid intake (5 liters/day) increases free cortisol excretion without an
increase in urine 17-hydroxycorticosteroids. Thus, mild to moderate
increases in urine cortisol excretion may not indicate hypercortisolism in individuals who have a high fluid intake and urine volume.
(J Clin Endocrinol Metab 83: 682– 684, 1998)
T
ditions of normal and increased fluid intake in normal
volunteers.
HE URINE free cortisol is one of the principal screening
tests for hypercortisolism (1). Although urine free cortisol is sometimes conceived as representing all cortisol that
is filtered at the glomerulus, the amount of filtered cortisol
actually greatly exceeds the amount of free cortisol measured
in the urine. For example, a glomerular filtration rate of 180
liters/1.73 m2 per day (2) and a mean plasma free cortisol
concentration of approximately 5 mg/L (3, 4) would produce
a filtered cortisol load of 900 mg/1.73 m2 per day. This
amount is approximately 20 times the upper limit of the
normal urine free cortisol, as determined by highly specific
assays on chromatographed samples (5, 6). Thus, most of the
filtered cortisol must either be metabolized or reabsorbed
before it is excreted. Two recognized renal metabolic reactions are catalyzed by 11b-hydroxysteroid dehydrogenase,
the site of the defect in the syndrome of apparent mineralocorticoid excess (7, 8), and 20-ketosteroid reductase, an enzyme implicated in rare patients with Cushing syndrome
and normal or low urine free cortisol (9, 10).
The fact that most filtered cortisol is metabolized or reabsorbed suggested the hypothesis that increased fluid intake and the resulting increase in urine volume might reduce
the fraction of filtered cortisol that is metabolized or reabsorbed and thus increase the urine free cortisol. To test this
hypothesis we measured urine cortisol excretion under con-
Subjects and Methods
Subjects
We studied six normal volunteers (three females, three males) with
normal body mass index (mean 23.2 6 3.6) and ages from 22– 45 yr. The
study protocol was approved by the institutional review board of the
National Institute of Child Health and Human Development, and informed consent was obtained from each subject. Volunteers were assigned, according to a randomized, cross-over design to 5 days of normal
fluid intake and 5 days of high fluid intake. A 2-day washout period of
normal fluid intake separated the two arms. During the 5 days of high
fluid intake, patients were asked to drink 5 liters of fluid per day, with
water comprising most of the excess fluid above their normal fluid
intake.
Urine assays
Creatinine in each 24-h urine specimen was measured at the National
Institutes of Health Clinical Center. Urine free cortisol was measured at
Hazleton Labs. (Vienna, VA) by a modification of a previously described
RIA (11). The cortisol antibody was raised against cortisol-3-carboxymethyloxime conjugated to BSA. The antibody cross-reactivity (relative to
100% for cortisol) was 37% for 11-deoxycortisol; 25% for cortisone; 3%
for corticosterone; 0.8% for 17-hydroxyprogesterone; and , 0.02% for
progesterone, pregnenolone, dehydroepiandrosterone, and testosterone. After dichloromethane extraction, samples were chromatographed
on celite columns (with elution of the cortisol fraction by 40% ethyl
acetate) before assay. The intra - and interassay coefficients of variation
were 9.5% and 13%, respectively.
An aliquot from each urine collection was also analyzed for 17hydroxycorticosteroids by a modification of the Porter-Silber method
(10). The intra - and interassay coefficients of variation were 5.9% and
7–14%, respectively. All 10 samples from each subject were measured in
the same urine free cortisol or 17-hydroxycorticosteroid assay to avoid
interassay variation. Additionally, during the period of high fluid intake,
Received April 7, 1997. Revision received August 12, 1997. Accepted
October 29, 1997.
Address all correspondence and requests for reprints to: M. Veronica
Mericq, National Institutes of Child Health and Human Development,
Developmental Endocrinology Branch, SDE, Building 10, Room 10N262,
10 Center Drive, MSC 1862, Bethesda, Maryland 20892-1862. E-mail:
[email protected].
* Commissioned officer in the United States Public Health Service.
682
COMMENTS
683
the volume of the urine aliquots was increased 3-fold to assay a similar
fraction of the total 24-h collection for each study period.
significantly [1.51 6 0.48 vs. 1.45 6 0.37 g/day, P 5 not
significant (NS)].
Statistical analysis
Urine free cortisol and 17-hydroxycorticosteroids
Comparisons between the normal and high-fluid periods were made
with the two-tailed paired Student’s t test. The frequency of elevated
urine free cortisol results for each study period was compared by the
x-square test. Unless otherwise stated, data are presented as mean 6 sd.
High fluid intake caused a significant increase in urine free
cortisol (126 6 33 vs. 77 6 18 mg/day, P , 0.005, Fig. 1) but
not in urine 17-hydroxycorticosteroids (5.3 6 1.5 vs. 5.0 6 1.7
mg/day, P 5 NS). The urine 17-hydroxycorticosteroid excretion per gram of creatinine excretion was nearly identical
between the two arms [3.49 6 0.51 (high fluid ) vs. 3.45 6 0.57
mg/g creatinine per day, P 5 NS). Additionally, the frequency of elevated urine free cortisol levels (normal range,
9 –95 mg/day) was significantly greater during the period of
high fluid intake (23/30 vs. 6/30, P , 0.005).
Results
Urine volume and creatinine
As expected, urine volume was significantly greater during the 5 days of high fluid intake than during the period of
normal intake (3800 6 1033 mL vs. 1070 6 376 mL, P , 0.005,
Fig. 1). By contrast, urine creatinine excretion did not change
Discussion
FIG. 1. Effect of high fluid intake on urine volume, urine free cortisol,
urine 17-hydroxycorticosteroids (17OHCS), and urine creatinine excretion in healthy normal volunteers. Each symbol represents data
from one of six subjects, and each data point shown is mean of five
consecutive daily measurements for that subject during each phase of
study. Shaded area, Normal range (9 –95 mg/day) for urine cortisol
excretion. Vertical bars, Mean 6 SEM for data from all six subjects.
* P , 0.005, high fluid intake vs. normal fluid intake.
We observed significantly higher urine cortisol excretion
during high fluid intake than during normal fluid intake.
This higher cortisol excretion occurred without an increase
in urine 17-hydroxycorticosteroid excretion, indicating that
the rise in urine cortisol did not appear to result from activation of the hypothalamic-pituitary-adrenal axis. Although
this study may lack sufficient power to exclude a real increase in 17-hydroxycorticosteroid excretion, our observations seem more consistent with the hypothesis that increased fluid intake decreases the renal metabolism or
reabsorption of filtered cortisol.
Like many urine free cortisol assays in current use, the RIA
used in this study lacks absolute specificity for cortisol and
thus yields values approximately twice as great as highly
specific methods (5, 6). The cause of such higher values is
cross-reaction of closely related cortisol precursors and metabolites (5, 6). Because these cross-reacting steroids (and
hence their fractional renal reabsorption and metabolism)
may differ among assays, it cannot be assumed that the effect
of high fluid intake will be precisely the same for different
assays. Indeed, the effect of fluid load could be on these
metabolites rather than on cortisol itself, and it would be of
interest to examine the effects of fluid loading using more
specific assay methods for urine cortisol.
Despite the widespread clinical use of urine free cortisol
measurements for the diagnosis of hypercortisolism, there
have been few clinical studies of the effect of fluid intake on
urine cortisol excretion. In 1959, Schedl and colleagues (12),
measuring urine cortisol by fluorometry and by double isotope dilution, stated that there was no correlation between
cortisol excretion and urine flow, but did not provide the
data for this conclusion. In 1974, Baum and colleagues (13)
published an abstract stating that water loading increased
the urine free cortisol and that water deprivation did the
reverse. To our knowledge, a full account of this study was
never published. Bertrand et al. (14) analyzed the factors
affecting the spontaneous excretion of free cortisol in 203
overnight urine collections from 7- to 18-yr-old school children. In these overnight urine collections (median volume
235 mL), total cortisol excretion was strongly related to urine
volume (P , 0.0001), with an increase of 24% (95% confidence limits: 17–32%) with each 100-mL increase in urine
volume. Thus, the limited previous data are consistent with
684
JCE & M • 1998
Vol 83 • No 2
COMMENTS
our observation that increased fluid intake increases the excretion of free cortisol.
Our observations do not address the relative contribution
of metabolism vs. reabsorption in explaining the discrepancy
between the amount of filtered and excreted cortisol. Although earlier studies made the assumption that reabsorption accounted for essentially all of the discrepancy (12, 15,
16), more recent studies have demonstrated renal metabolism of cortisol both by 20-ketosteroid reductase (9) and by
11b-hydroxysteroid dehydrogenase (17). Thus, the increase
in urine cortisol excretion during high fluid intake may reflect decreased renal metabolism of cortisol as well as decreased cortisol reabsorption.
Our study did not formally address the effect of more
modest urine volume increases such as would be encountered more commonly in clinical practice. However, the six
individual 24-h urine collections with the lowest volume
(3–3.5 liters) during high fluid intake had a mean 6 sd urine
free cortisol of 107 6 27 mg/day, with five of the six values
above the normal limit of 95 mg/day. Thus, it appears that
more modest increases in urine volume also produce proportional increases in urine free cortisol.
The clinical significance of these observations is that 77%
of the 24-h urine samples collected from normal volunteers
during high fluid intake had urine free cortisol levels that
were above the upper limit of the normal range. Thus, it
seems likely that patients with high fluid intake, such as
patients with psychogenic polydipsia or diabetes insipidus,
may also be at greater risk of excreting increased levels of
cortisol during periods of high fluid intake and urine volume. If this hypothesis is correct, mild to moderate increases
in urine cortisol in such patients should be confirmed by an
alternate method, such as the overnight dexamethasone suppression test, diurnal cortisol levels in plasma, or urine 17hydroxycorticosteroid excretion before concluding that the
increased urine cortisol level is in fact caused by increased
cortisol secretion (18).
References
1. Flack MR, Oldfield EH, Cutler GB, et al. 1992 Urine free cortisol in the
high-dose dexamethasone suppression test for the differential diagnosis of
Cushing syndrome. Ann Intern Med. 116:211–217.
2. Brenner BM, Rector Jr FC (eds). 1996 The kidney. 5th ed. Philadelphia: Saunders; 286 –333.
3. Ballard PL. 1979 Delivery and transport of glucocorticoids to target cells.
Chapter 2. In: Baxter JD, Rousseau GG, eds. Glucocorticoid hormone action.
New York: Springer-Verlag; 25– 49.
4. Dunn JF, Nisula BC, Rodbard D. 1981 Transport of steroid hormones:
binding of 21 endogenous steroids to both testosterone-binding globulin
and corticosteroid binding globulin in human plasma. J Clin Endocrinol
Metab. 2:103–123.
5. Murphy BEP, Okouneff LM, Klein GP, et al. 1981 Lack of specificity of cortisol
determinations in human urine. J Clin Endocrinol Metab. 53:91–99.
6. Lin CL, Wu TJ, Machacek DA, et al. 1997 Urine free cortisol and cortisone
determined by high performance liquid chromatography in the diagnosis of
Cushing syndrome. J Clin Endocrinol Metab. 82:151–155.
7. Ulick S, Ramirez LC, New MI. 1977 An abnormality in steroid reductive
metabolism in a hypertensive syndrome. J Clin Endocrinol Metab. 44:799 – 802.
8. New MI. 1994 The prismatic case of apparent mineralocorticoid excess. J Clin
Endocrinol Metab. 79:1–3.
9. Schoneshofer M, Weber B, Nigam S. 1983 Increased urinary excretion of free
20a- and 20b-dihydrocortisol in a hypercortisolemic but hypocortisoluric patient with Cushing’s disease. Clin Chem. 29:285.
10. Porter CC, Silber RH. 1950 A quantitative color reaction for cortisone and
related 17,21 dihydroxy-20-ketosteroids. J Biol Chem. 185:201–206.
11. Ruder HJ, Guy RL, Lipsett MB. 1972 A radioimmunoassay for cortisol in
plasma and urine. J Clin Endocrinol Metab. 35:219 –224.
12. Schedl HP, Chen PS, Greene G, Redd D. 1959 The renal clearance of plasma
cortisol. J Clin Endocrinol Metab. 19:1223–1229.
13. Baum CK, Davison MJ, Landon J. 1974 Urinary free cortisol excretion by
normal subjects. J Endocrinol. 63:47– 48.
14. Bertrand RV, Rudd TB, Weller PH, Day AJ. 1987 Free cortisol and creatinine
in urine of healthy children. Clin Chem. 33:2047–2051.
15. Beisel WR, Cos JJ, Horton R, Chao PY, Forsham PH. 1964 Physiology of
urinary cortisol excretion. J Clin Endocrinol. 24:887– 893.
16. Lindholm J. 1973 The renal excretion of cortisol. J Clin Lab Invest. 31:115–118.
17. Edwards CR, Stewart PM, Burt D, et al. 1988 Localization of 11 b-hydroxysteroid dehydrogenase-tissue-specific protector of the mineralocorticoid receptor. Lancet. 2:986 –989.
18. Nieman LK, Cutler Jr GB. 1995 Cushing’s syndrome. In: De Groot LJ, ed.
Endocrinology. 3rd ed. Philadelphia: Saunders; 1741–1769.