ORIGINAL ARTICLE

E n d o c r i n e C a r e

Control of Childhood Congenital Adrenal Hyperplasia

and Sleep Activity and Quality with Morning or
Evening Glucocorticoid Therapy

Alina German, Suheir Suraiya, Yardena Tenenbaum-Rakover, Ilana Koren, Giora Pillar,
and Ze’ev Hochberg
Pediatric Endocrinology (A.G., Z.H.) and Sleep Laboratory (S.S., G.P.), Rambam Medical Center, Haifa 31096, Israel; Pediatric Endocrine
Unit (A.G., I.K.), Clalit Health Maintenance Organization, Haifa 35013, Israel; Ha’Emek Medical Center (Y.T.-R.), Afula 18101, Israel;
and the Faculty of Medicine, Technion–Israel Institute of Technology (Y.T.-R., G.P., Z.H.), Haifa 32000, Israel

Context: Traditionally, hydrocortisone (HC) replacement therapy in congenital adrenal hyperplasia

(CAH) is given by three daily doses, albeit not necessarily of equal quantity. Although a higher dose
in the morning better imitates the physiological diurnal variation, a late-night higher dose was
suggested to better suppress early morning hypothalamic-pituitary-adrenal axis peak activity. Yet,
increased night cortisol has been claimed to be associated with sleep disturbances and insomnia.

Objective: Our objective was to evaluate evening vs. morning high-HC dose with respect to disease
control, sleep pattern, and daytime activity in children with CAH.

Design: An open-label, cross-over, randomized trial of 15 children with classical CAH was per-
formed. Patients were randomized to receive 50% of the daily HC in the morning or evening for
2 wk; the other two doses included 25% of the daily dose each.

Outcome Measures: Disease control was assessed by 0800-h 17-hydroxyprogesterone, testoster-

one, androstenedione, and dehydroepiandrosterone sulfate on the last day of each treatment
schedule. Sleep and daytime activity were assessed by a 7-d actigraph.

Results: Basal morning androstenedione, 17-hydroxyprogesterone, dehydroepiandrosterone sul-

fate, and testosterone levels during the high-morning and high-evening HC treatment schedules
were comparable. There were no significant differences in sleep or daytime activity.

Conclusions: With respect to disease control, sleep quality and daytime activity were not affected
by treatment schedules. We recommend the high-morning dose schedule in replacement therapy
of children with CAH. (J Clin Endocrinol Metab 93: 4707– 4710, 2008)

T he main goal in congenital adrenal hyperplasia (CAH) ther-

apy is to approach a physiological replacement of cortisol
deficiency while suppressing adrenal androgen overproduction.
Traditionally, treatment is given by three daily doses, albeit not
necessarily of equal quantity. Although a higher dose in the
morning better imitates the physiological diurnal variation, a
late-night higher dose was suggested to better suppress early
morning hypothalamic-pituitary-adrenal (HPA) axis peak
activity.
It has been claimed that increased evening cortisol levels are
associated with sleep disturbances and insomnia (1) because a
positive correlation was demonstrated between evening cortisol
levels and the number of nocturnal awakenings in normal sub-
jects (1, 2). The circadian clock, rather than sleep, drives the daily
changes in cortisol levels (3). In normal individuals there is a
marked circadian rhythm in cortisol release, with the lowest lev-
els occurring shortly after 2400 h, and increasing between 0200
and 0300 h to peak at around 0600 – 0800 h after waking.

0021-972X/08/$15.00/0 Abbreviations: CAH, Congenital adrenal hyperplasia; CV, coefficient of variation; DHEAS,
Printed in U.S.A. dehydroepiandrosterone sulfate; HC, hydrocortisone; 17OHP, 17-hydroxyprogesterone;
HPA, hypothalamic-pituitary-adrenal axis; SWS, slow-wave sleep.
Copyright © 2008 by The Endocrine Society
doi: 10.1210/jc.2008-0519 Received March 4, 2008. Accepted August 28, 2008.
First Published Online September 9, 2008

J Clin Endocrinol Metab, December 2008, 93(12):4707– 4710 jcem.endojournals.org 4707

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 November 2015. at 06:35 For personal use only. No other uses without permission. . All rights reserved.
4708 German et al. Sleep in CAH
Hormones of the HPA axis possibly play a significant role in
determining entry into and duration of different sleep and
activity stages. In normal individuals the HPA axis is most
active in the second half of the night, with as much as 70% of
the total nighttime cortisol secretion occurring during that
time, positively correlating with rapid eye movement sleep
duration (4), whereas declining plasma cortisol levels are as-
sociated with brain lateralization, slow-wave sleep (SWS), or
deep sleep (2, 5).
The present study was designed to evaluate evening vs. morn-
ing administration of a higher hydrocortisone (HC) dose with
respect to disease control, sleep pattern, and daytime activity in
children with CAH in an open-label, randomized, cross-over
design. This was evaluated with a sleep log as well as a wrist
actigraph, which was shown to provide a reasonably accurate
estimation of sleep and wakefulness (6).
Patients and Methods
Study design
This was an open-label, cross-over, randomized trial, designed to
compare disease control, sleep quality, and daytime activity with a high-
morning or high-evening HC dose in children with CAH. Patients were
randomized in two groups for their initial treatment schedule: a high-
morning dose, when 50% of the daily HC was taken in the morning; and
a high-evening dose, when 50% was taken at bedtime. The other two
doses included 25% of the daily dose each. The schedule was standard-
ized to 0700 – 0800, 1300 –1400, and 2100 –2200 h according to pa-
tients’ age. Duration of the study was 4 wk: 2 wk for each treatment
schedule. Patients received standard replacement therapy with an oral
HC dose that was identical to each subject’s prestudy therapy, ranging
from 13.5–15.3 mg/m2 given three times daily; 9␣-fludrocortisone was
given once a day with the morning HC tablets, and median plasma renin
activity level was 3.3 ng/ml䡠24 h (range 2.87–3.55). The protocol was
approved by the Rambam Medical Center Ethics committee, conducted
in accordance with the Declaration of Helsinki, and a written, informed
consent was signed by a parent.
Patients
There were 18 children (10 males and eight females; median age 10
yr, range 6 –17) with normal circadian rhythm and classical CAH due
to 21-hydroxylase deficiency (n ⫽ 14) or 11-hydroxylase deficiency
(n ⫽ 1) with reasonable disease control 关as determined by mean 17-
hydroxyprogesterone (17OHP) levels in the previous year ⬍30 nmol/
liter兴 enrolled in the study. Five children had received a high-morning
dose, six a high-evening dose, and four received three equal daily
doses. Three patients from the high-evening dose group were excluded
during the study for compliance reasons, and, therefore, the group
sizes are not balanced. Exclusion criteria included known sleep, be-
havioral, or movement disturbances. Patients’ clinical characteristics
are presented in Table 1.
Clinical and laboratory assessment
Assessment of disease control was done by measuring before the
morning dose 0800-h levels of 17OHP, testosterone, androstenedione,
and dehydroepiandrosterone sulfate (DHEAS) on the last day of each
treatment schedule. Reference values were retrieved from Ref. 7. Sleep
and daytime activity were assessed by a 7-d actigraph, monitoring total
sleep time, number of arousals, sleep efficiency, sleep latency, and day-
time movements on the second week of each treatment schedule (8).
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 November 2015. at 06:35 For personal use only. No other uses without permission. . All rights reserved.
J Clin Endocrinol Metab, December 2008, 93(12):4707– 4710
TABLE 1. Clinical characteristic of the study’s 15 patients with
CAH
No. of males/females 9/6
Age (yr) 10 (7.5; 14.5)
No. of prepubertal/pubertal patients 9/6
BMI (kg/m²) 19.7 (14.8; 23.4)
No. of 21-hydroxylase/11-hydroxylase deficiency 14/1
HC dose (mg/m2䡠d) 14 (13.5; 15.3)
No. of initial HC dose (morning/evening) 9/6
Fludrocortisone dose (␮g/m2䡠d) 130 (72; 145.8)
Data shown are median (first; third quartiles). BMI, Body mass index.
Assays
17OHP was measured using the Coat-A-Count RIA (Coat-A-
Count; Siemens Healthcare Diagnostics, Los Angeles, CA) with a
sensitivity of 0.212 nmol/liter. The within-assay coefficients of vari-
ation (CVs) were 6.7 and 3.5% at serum concentrations 0.3 and 6.5
ng/ml, respectively, and the between-assay CVs were 11 and 8.5% at
serum concentrations of 0.35 and 6.1 ng/ml, respectively. Andro-
stenedione serum levels were measured using a Diagnostic Systems
Laboratories RIA kit (Webster, TX) with a sensitivity of 0.069 nmol/
liter. Intraassay CVs were 4.3 and 5.9% at serum concentrations of
0.69 and 6.9 ng/ml, respectively, and interassay CVs of 6.3 and 7.0%
at serum concentrations of 0.63 and 6.51 ng/ml, respectively. The
DHEAS level was measured using Siemens Healthcare Diagnostics
RIA (Immune 2000 analyzer system chemiluminescent enzyme im-
munoassay), with a sensitivity of 0.08 ␮mol/liter. The within-assay
CVs were 9.8 and 4.3% at serum concentrations of 52 and 659 ␮g/dl,
respectively. Testosterone levels were measured using Chemilumines-
cent immunoassay (Advia Centaur analyzer chemiluminescent immu-
noassay; Bayer Health Care, LLC, Fernwald, Germany), with a sen-
sitivity of 0.35 nmol/liter. Intraassay CVs were 6.2 and 2.3% at serum
concentrations 3.31 and 35.03 nmol/liter, respectively, and interassay
CVs of 4.4 and 1.4% at serum concentrations of 3.31 and 35.03
nmol/liter, respectively.
Actigraphy
Sleep quality and daytime activity were identified by continuous ac-
tigraph monitoring. The actigraph (Individual Monitoring Systems, Inc.,
Baltimore, MD) is a small watch-like device weighing 10 g that records
movement accelerations. It is worn all day and night (24 h except show-
ers) on the wrist of the nondominant hand. Using a validated accurate
algorithm to translate the presence of movement accelerations and score
awake or sleep (9, 10), the actigraph data were processed and scored for
the following variables: time in bed (the time from lights off until lights
on); total sleep time (the actual time the child really slept); total wake time
(wake after sleep onset, i.e. the time the child spent awake between falling
asleep and lights on); sleep onset latency (the time from reported lights
off till the time of falling asleep); sleep efficiency (the total sleep time as
a percentage of time in bed); arousal number; and activity index (intensity
of movement acceleration during a given time).
Statistical analyses
The primary endpoints for the study were sleep and activity variables:
the changes in activity index and the number of arousals per night. The
sample size was selected to allow for 80% power to detect a difference
of 20 U activity and two nocturnal arousals per night between the two
groups with SD at an ␣-level of 0.05.
Comparison of endocrine results used the nonparametric Wilcoxon
signed-ranks paired test, and sleep results, with their normal distribution,
used the Student’s t test. Therefore, results are given as median and
quartiles for biochemical parameters, and mean ⫾ SD for sleep
parameters.
J Clin Endocrinol Metab, December 2008, 93(12):4707– 4710 jcem.endojournals.org 4709

TABLE 2. Endocrine parameters in children receiving a high-morning or high-evening HC dose

Reference prepuberty High-morning High-evening Morning 关ⴚ兴

mean ⴞ 2 SD dose dose evening difference
17OHP (␮mol/liter) 0.09 –2.7 44 (16; 116) 33 (15; 76) 3.1 (⫺27; 13)
DHEAS (␮mol/liter) 0.2–2.0 0.2 (0.2; 0.6) 0.4 (0.2; 0.7) 0 (⫺0.3; 0)
Androstenedione (nmol/liter) 0.3–1.8 1.8 (1.0; 3.0) 1.9 (1.2; 6.5) ⫺0.2 (⫺3.6; 0)
Testosterone (nmol/liter) ⬍0.1– 0.4 0.7 (0.3; 2.3) 1.1 (0.6; 2.7) 0 (0.6; 2.7)
Hormonal concentrations were obtained on the last morning of each treatment schedule. Median (first; third quartiles). All differences are insignificant. Reference levels
for prepubertal children are adapted from Ref. 7.

Results Discussion

CAH control assessment Comparing high-morning or high-evening HC dose schedules in

Biochemical parameters in children receiving a high-morn- children with CAH, the present study shows that with respect to
ing or high-evening HC dose are presented in Table 2. Basal disease control, sleep quality, and daytime activity, the two
morning levels of 17OHP, DHEAS, androstenedione, and tes- schedules were comparable.
tosterone were comparable. The patient with 11-hydroxylase The randomized, cross-over design is one of the strong points
deficiency was excluded from 17-OHP comparison analysis. in this study. The short-term nature of this 4-wk trial allows for
Three pubertal (Tanner stages III-V) males were excluded conclusions of short-term parameters, such as sleep, wakeful-
from comparison analysis of testosterone and androstenedi- ness, and disease control. Conclusions cannot be drawn on long-
term glucocorticoid effects such as growth, bone maturation,
one levels. The table shows a wide range of differences be-
and body composition.
tween the evening and morning high-dose groups. Subjects
The study has a number of potential limitations, which
who started the study with a high-morning dose did not differ
have been considered. Defining primary endpoints of sleep
in their endocrine parameters from those starting with a high-
and activity resulted in a relative small sample size for endo-
evening dose, and the second regimen did not differ from the
crine analysis, which has been defined as secondary endpoints.
first regimen.
Although, the comparable endocrine results of the two groups
were unequivocal. Hormonal measurements were done at
Sleep quality and daytime activity 0800 h only, and as such do not reflect 24-h hormonal status.
Arousal number, total sleep time, sleep efficiency, and sleep It is realized that 2 wk may not be long enough to reflect
latency are presented in Table 3. There were no significant disease control.
differences in these sleep parameters between high-HC morn- In a study of oral HC bioavailability in CAH patients, it
was demonstrated that maximal cortisol concentrations are
ing and high-evening treatment schedules. Daytime 24-h ac-
apparent within 1–2 h, and levels become undetectable within
tivity was divided into eight 3-h periods starting at 0800 h
7 h after a morning oral dose and 9 h after an evening oral HC
(Fig. 1). There were no differences in daytime activity indices
administration (11, 12). HC clearance in the evening was 26%
between the two treatment schedules. Subjects who started the
slower than clearance in the morning, probably secondary to
study with a high-morning dose did not differ in their sleep
circadian variation in corticosteroid-binding globulin binding
parameters from those starting with a high-evening dose.
capacity and ACTH concentrations (13, 14). Assuming this
pharmacokinetics of cortisol at night, a high-bedtime HC
dose, taken at 2100 –2200 h, would give peak levels between
TABLE 3. Sleep parameters of morning and evening
treatment schedules

Normal
Morning Evening childhood
dose dose ranges
Arousal no./night 4.7 ⫾ 2.8 4.4 ⫾ 1.7 0–6
Total sleep time (min/night) 462 ⫾ 50 531 ⫾ 46 360 – 600
Sleep efficiency (%) 85.8 ⫾ 7.6 87 ⫾ 6.4 92–100
Activity index 24 h 113 ⫾ 25 109 ⫾ 20
Sleep onset latency (min) 23 ⫾ 18 24 ⫾ 17 0 –20
Values are shown as mean ⫾ SD. No reference ranges are determined for
24-h activity index. All differences are insignificant. Activity index, Intensity
of movement acceleration during a given time; Sleep efficiency, total sleep
time as a percentage of time in bed; Sleep onset latency, time from reported FIG. 1. Comparison of 24-h activity index during high-morning (open bars) or
lights off till the time of falling asleep; Total sleep time, actual time the child high-evening (solid bars) HC dose treatment. Daytime 24-h activity was divided
really slept. into eight 3-h periods. Error bars represent SD values.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 November 2015. at 06:35 For personal use only. No other uses without permission. . All rights reserved.
4710 German et al. Sleep in CAH
2200 and 2400 h, and provide glucocorticoid coverage for a
greater length of time than an equal dose administered at
0700 – 0800 h and reach an undetectable level at early morn-
ing. It was suggested that this would provide disease control
that is comparable to the high-morning dose schedule (11, 12).
That 17OHP values did not differ between the two schedules
may be due to its strong responsiveness to stress. Indeed, the
variability within and among patients was wide. To verify
whether this indicates that patients went out of control after the
change, we compared the groups and found that subjects who
started the study with a high-morning dose did not differ in their
endocrine parameters from those starting with a high-evening
dose, and the second regimen did not differ from the first
regimen.
It was previously shown that the 24-h 17OHP concentrations
demonstrate circadian variation, with peak values observed be-
tween 0400 and 0900 h, and that 17OHP concentrations de-
crease gradually in response to the increase in cortisol concen-
trations during the day, but remain low during the night, despite
low cortisol concentrations between 1600 and 2200 h (11).
These and the present results suggest that administration of oral
HC in the evening may be required for replacement purposes but
cannot be expected to contribute significantly toward HPA axis
suppression.
Rejecting the study hypothesis, and in contrast to reports on
glucocorticoid effect on adult wakefulness (15), the total sleep
time, sleep efficiency, arousal number, and 24-h activity index
were not different among the treatment schedules. Children may
have deeper sleep that overcomes an apparent glucocorticoid
effect in the dose used here. Indeed, it is well documented that
sleep in children is deeper than in adults, and it has even recently
been suggested that cortical electroencephalogram arousal pat-
terns are not sensitive enough to detect arousals from sleep in
children due to the deep nature of their sleep. Instead, outputs of
the autonomic nervous system were suggested as a better alter-
native to detect arousals from sleep (16, 17). These results are
compatible with a previously shown opposite dose-dependent
effect of exogenous glucocorticoids on electroencephalogram
(18). It was shown that the effect of exogenous and endogenous
cortisol on SWS depends on optimal cortisol levels for maximal,
nocturnal, mineralocorticoid, receptor-mediated CRH suppres-
sion. Very high levels of cortisol and stress, by the glucocorticoid
receptor, exert excitatory feedback on CRH, characterized by
decreasing SWS and increasing wakefulness (15, 19).
In conclusion, and in agreement with a previous report (12),
whereas nighttime administration does not interfere with sleep
and activity, there is no advantage in giving HC high dose in the
evening. High-dose HC in the evening and high-dose HC in the
morning were found to be equivalent with respect to disease
control, sleep pattern, and daytime activity.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 07 November 2015. at 06:35 For personal use only. No other uses without permission. . All rights reserved.
J Clin Endocrinol Metab, December 2008, 93(12):4707– 4710
Acknowledgments
Address all correspondence and requests for reprints to: Ze’ev Hochberg,
M.D., Ph.D., Meyer Children’s Hospital, Rambam Medical Center,
Haifa 31096, Israel. E-mail: z_hochberg@rambam.health.gov.il.
Disclosure Statement: The authors have nothing to declare.
References
1. Chrousos GA, Kattah JC, Beck RW, Cleary PA 1993 Side effects of glucocor-
ticoid treatment. Experience of the Optic Neuritis Treatment Trial. JAMA
269:2110 –2112
2. Vgontzas AN, Chrousos GP 2002 Sleep, the hypothalamic-pituitary-adrenal
axis, and cytokines: multiple interactions and disturbances in sleep disorders.
Endocrinol Metab Clin North Am 31:15–36
3. Lavie P, Pillar G, Malhotr A 2002 Sleep disorders: diagnosis, management and
treatment. London: Martin Dunitz
4. Garcia-Borreguero D, Wehr TA, Larrosa O, Granizo JJ, Hardwick D, Chrousos
GP, Friedman TC 2000 Glucocorticoid replacement is permissive for rapid eye
movement sleep and sleep consolidation in patients with adrenal insufficiency.
J Clin Endocrinol Metab 85:4201– 4206
5. Tirosh E, Rod R, Cohen A, Hochberg Z 1993 Congenital adrenal hyperplasia
and cerebral lateralizations. Pediatr Neurol 9:198 –201
6. Vallieres A, Morin CM 2003 Actigraphy in the assessment of insomnia. Sleep
26:902–906
7. Diamond FJ, Bercu BB 2004 Normative laboratory results. In: Pescovitz OH,
Eugster EA, eds. Pediatric endocrinology: mechanisms, manifestations, and
management. Philadelphia: Lippincott Williams & Wilkins; 780 – 823
8. Hedner J, Pillar G, Pittman SD, Zou D, Grote L, White DP 2004 A novel
adaptive wrist actigraphy algorithm for sleep-wake assessment in sleep apnea
patients. Sleep 27:1560 –1566
9. Acebo C, Sadeh A, Seifer R, Tzischinsky O, Wolfson AR, Hafer A, Carskadon
MA 1999 Estimating sleep patterns with activity monitoring in children and
adolescents: how many nights are necessary for reliable measures? Sleep 22:
95–103
10. Sadeh A, Lavie P, Scher A, Tirosh E, Epstein R 1991 Actigraphic home-mon-
itoring sleep-disturbed and control infants and young children: a new method
for pediatric assessment of sleep-wake patterns. Pediatrics 87:494 – 499
11. Charmandari E, Johnston A, Brook CG, Hindmarsh PC 2001 Bioavailability
of oral hydrocortisone in patients with congenital adrenal hyperplasia due to
21-hydroxylase deficiency. J Endocrinol 169:65–70
12. Charmandari E, Matthews DR, Johnston A, Brook CG, Hindmarsh PC 2001
Serum cortisol and 17-hydroxyprogesterone interrelation in classic 21-hy-
droxylase deficiency: is current replacement therapy satisfactory? J Clin En-
docrinol Metab 86:4679 – 4685
13. Angeli A, Frajria R, De Paoli R, Fonzo D, Ceresa F 1978 Diurnal variation of
prednisolone binding to serum corticosteroid-binding globulin in man. Clin
Pharmacol Ther 23:47–53
14. Hsu BR, Kuhn RW 1988 The role of the adrenal in generating the diurnal
variation in circulating levels of corticosteroid-binding globulin in the rat.
Endocrinology 122:421– 426
15. Buckley TM, Schatzberg AF 2005 On the interactions of the hypothalamic-
pituitary-adrenal (HPA) axis and sleep: normal HPA axis activity and circadian
rhythm, exemplary sleep disorders. J Clin Endocrinol Metab 90:3106 –3114
16. Etzioni T SS, Pillar G 2007 Sleep disordered breathing events are associated with
autonomic activations in children with OSA. Sleep 30:A68 –A69 (Abstract)
17. O’Brien LM, Gozal D 2007 Potential usefulness of noninvasive autonomic
monitoring in recognition of arousals in normal healthy children. J Clin Sleep
Med 3:41– 47
18. Follenius M, Brandenberger G, Bandesapt JJ, Libert JP, Ehrhart J 1992 Noc-
turnal cortisol release in relation to sleep structure. Sleep 15:21–27
19. Hochberg Z, Pacak K, Chrousos GP 2003 Endocrine withdrawal syndromes.
Endocr Rev 24:523–538