Downloaded from http://jnnp.bmj.com/ on June 18, 2017 - Published by group.bmj.com
J. Neurol. Neurosurg. Psychiat., 1971, 34, 432-435
Cortisol in the cerebrospinal fluid of patients
suffering from affective disorders
ALEC COPPEN, B. W. L. BROOKSBANK, R. NOGUERA, AND D. A. WILSON
From the Medical Research Council Neuropsychiatric Research Unit, Carshalton, and at West Park Hospital,
Epsom, Surrey
Cortisol concentration was estimated by a competitive protein-binding method in the
lumbar cerebrospinal fluid (CSF) of patients suffering from depression or from mania, and compared
with the CSF cortisol concentration of neurological control patients not suffering from an affective
disorder. There were no significant differences between the groups.
SUMMARY
There have been numerous investigations into
adrenocortical function in affective disorders using
either 1 7-hydroxycorticosteroid excretion rate,
measurement of plasma cortisol, or estimates of the
secretion rate of cortisol by isotope dilution techniques (see reviews by Sachar, 1967; Coppen, 1970).
Reports have been conflicting, but on the whole
the evidence has suggested that there is a moderate
increase in adrenocortical activity in a proportion
of patients admitted to hospital suffering from
depression, although this decreases as the patient
adapts to his new environment. The increased
adrenocortical activity decreases in most, though not
in all cases, with clinical recovery. Most investigations of adrenocortical activity have been concerned
with peripheral measures as opposed to its estimation in the central nervous system where it is possible
that it could influence the formation of 5-hydroxytryptamine (Curzon, 1969; Lapin and Oxenkrug,
1969), which is known to be abnormal in the affective
disorders (Coppen, 1968).
One approach to the study of cortisol in the
central nervous system is to measure its concentration in cerebrospinal fluid (CSF). CSF contains
unbound cortisol which is probably in equilibrium
with the unbound cortisol in blood (Murphy, Cosgrove, Mcllquham, and Pattee, 1967). McClure
(1969) in a small series of five depressed patients
found them to have considerably lower levels than
those found in five control subjects.
During an investigation into indoleamine metabolism in the affective disorders, samples of lumbar
CSF were obtained from patients suffering from
depression, or mania, and from neurological patients
who were not affectively disturbed who acted as a
control group. The opportunity was taken to
examine many of these samples of CSF for their
concentration of cortisol.
METHODS
PATIENTS Samples of CSF were obtained from 23
patients suffering from depression, 16 manic patients,
and 18 control patients with a variety of neurological
conditions but who did not have any psychiatric illness.
Severity of depression was measured by the Hamilton
Rating Scale (Hamilton, 1967) and the self-rating inventory devised by Beck, Ward, Mendelson, Mock, and
Erbaugh (1961). Severity of mania was assessed on a
five-point (0-4) global rating scale. CSF was obtained
from a lumbar tap a few days after admission (mean 2-2
days after admission), and care was taken to collect the
first 11 ml. of fluid. The CSF was then stored at - 15°C
until assay.
All patients suffering from affective disorders were
taken off their drugs one to two days before the test and
were sedated at night with chloral hydrate. Nine patients
suffering from depression had not received any drugs
for their psychiatric condition but their mean cortisol
level did not differ from the rest of the patients who had
been treated with a variety of antidepressant drugs before
admission.
ESTIMATION OF CORTISOL IN CSF Cortisol concentration
in CSF specimens was estimated by competitive protein-
binding radioassay (Murphy, 1967) using modifications
of the methods described by Murphy et al. (1967) for
CSF and by Beardwell, Burke, and Cope (1968) for urine.
The principal modifications were the addition of an
internal standard and the use of Fuller's earth in place
of Florisil for adsorption of unbound cortisol. The procedure was as follows:
To 1 ml. samples of CSF, duplicated if possible, was
added an accurately measured quantity of 1,2-3H cortisol
(Radiochemical Centre, Amersham, U.K., 2 c/mM)
(01 ml. ethanol containing about 10 m,uc) and 7 5 ml
432
Downloaded from http://jnnp.bmj.com/ on June 18, 2017 - Published by group.bmj.com
Cortisol in the cerebrospinal fluid of patients suffering from affective disorders
dichloromethane (Merck, 'pro analysi'). The mixture
was rotated on a Matburn mixer for 20 min. The aqueous
phase was removed and 5-0 ml. of the dichloromethane
extract taken to dryness in an assay tube under a nitrogen
stream at 35°C. The residue was taken up in 10 ml. of
2 Y ethanol in 0 9 % saline, and 0-2 ml. was removed for
determination of recovery of 3H. The percent recovery
ranged from 66 to 100.
Corticosteroid-binding globulin (CBG) isotope solution was prepared as described by Beardwell et al. (1968)
but utilizing human late pregnancy plasma. One millilitre of CBG isotope solution was added to the remaining
0-8 ml. of the extract and to 0-8 ml. 2o% (v/v) ethanol in
saline containing 0, 5, 10, and 15 mtsg cortisol; the
standards were each run in duplicate. The assay tubes
were held at 45°C for 5 min, then at 5°C for 10 min.
Fuller's earth (B.D.H. 'for adsorption', 30 mg) was
added to each tube, which was shaken for 2 min, centrifuged for 2 min at 5°C, and then returned to the 5°C
water-bath for a further 10 min. A sample (0 5 ml.) of
the supernatant was taken for scintillation counting. The
scintillator mixture consisted of 13 ml. toluene-ethoxyethanol (7:6) containing 044% 2,5-diphenyloxazole
(PPO) and 0-01 % 1 ,4-bis-2(4-methyl-5-phenyloxazolyl)benzene(Dimethyl-POPOP). Calculations were based on
a minimum of 20,000 counts in a 3H channel, using a
Packard Tricarb Model 3375 liquid scintillation spectrometer.
The standard curve, set up with each batch of samples,
linear over the range 0 to 15 mrng cortisol and the
coefficient of variation of the standard counts per min
was +1-67%. The mean percent bound 3H-cortisol was
68-9 ± 3-4 (SD), with a range of 63-0-73-1.
The concentration of cortisol in CSF was calculated
from the standard curve for the day, after correcting for
cortisol added as internal standard before extraction, the
percentage of this bound to protein and the losses on
extraction. In the range 0 to 6 m,4g/ml. CSF, the SD of
the difference between duplicates was ±126 miLg/ml.
As the greater part of each CSF sample was required for
was
433
other assays, the number of samples that could be
assayed for cortisol in duplicate was only 13. It was,
therefore, not possible to determine whether the accuracy
of the method differed significantly over the range of
observed values.
RESULTS
Table 1 shows the data obtained for the concentration of cortisol in CSF for each group, subdivided
according to sex, together with the ages of each
group and the correlations between CSF cortisol
and age. It is apparent that there were no significant
differences either between controls, depressives, and
manics or between sexes. Correlations with age were
significant only for the male controls (negative) and
male manics (positive). In view of the relatively high
methodological variance of the figures for CSF
cortisol concentration, no attempt at interpretation
of these unexpected correlations with age seems
warranted: in the male manic group the significance
vanishes if one aberrant value (9.9 m,ug/ml.) is
eliminated.
The Figure depicts the concentration of CSF
cortisol plotted against time of day of lumbar puncture in the subjects for whom this time was known.
It is clear that there is no diurnal trend discernible
in the whole population of figures, and that there is
insufficient information to tell whether diurnal
swings might differ between the clinical groups. No
patient was tested more than once on the same day.
There was also no significant correlation between
CSF cortisol concentration and severity of depression as measured by the Hamilton Scale (n = 13,
r, = +0 512) or by the Beck Scale (n = 20, r5 =
+0 333). In measuring the severity of mania in the
TABLE 1
CSF CORTISOI CONCENTRATION IN CONTROL PATIENTS AND IN PATIENTS SIuFFERING FROM DEPRESSION OR MANIA, AND
SPEARMAN RANK CORRELATIONS BETWEEN CSF CORTISOL CONCENTRATION AND AGE
Group
Number
Lumbar CSF cortisol
(mpsg/ml.)
Age (yr.)
Correlation CSF cortisol
mean
SD
mean
SD
rS
13
2-93
±2-09
±11-6
3-12
±2-62
46-5
5
±13-0
±12-1
-0-687
0-500
-0-442
vs. age, P
Controls
Males
Females
All subjects
Depressives
Males
Females
All patients
18
2.981
±2-17
51-6
47-9
11
12
23
2-89
3-68
3-301
±1-39
±2-35
±1-95
56-4
52-3
54-3
± 6-1
±10-3
± 8-6
0300
0-523
0-407
Manics
Males
Females
All patients
11
5
3-38
1-96
±2-69
±1-72
2-941
±2-47
45-7
42-4
44-7
±10-4
±15-5
16
±11-8
0-659
-0-075
0-248
'Controls vs. depressives, NS.
Controls vs. manics, NS.
<0 01
NS
NS
NS
NS
NS
<005
NS
NS
Downloaded from http://jnnp.bmj.com/ on June 18, 2017 - Published by group.bmj.com
Alec Coppen, B. W. L. Brooksbank, R. Noguera, and D. A. Wilson
434
9
A wControts
8k
*a Depressives
* = Manics
A
0
7F
E
-
0
-
11:
cli
C-
6k
5F
43k
A
.
I
0
A
A
A
2
FIG. CSF cortisol concentrations and time of day of
lumbar puncture.
0
A
a
a
U
U
A
.
I
10am
2pm
Time
-12noon
4pm
manic patients there were only three scale points to
compare, and so a chi-square test was carried out on
a null hypothesis: this gave a non-significant value
of X2 = 1-5 (n = 16).
Finally, in the group of patients suffering from
depression the CSF cortisol concentration was found
not to be related to the length of the interval between
admission to hospital and the lumbar tap (Table 2).
TABLE 2
CSF CORTISOL CONCENTRATION (MaLG/ML.) AND INTERVAL
(DAYS) BETWEEN ADMISSION TO HOSPITAL AND LUMBAR
PuJNCTURE
Days
Number of patients
Mean cortisol concentration
0
5
1
6
3-34
3-42
2
3
3-13
3-+4
3
2-97
5+
6
3 40
DISCUSSION
The control figures we have obtained (2-98 ± 2-17
(SD) m,ug/ml.) for cortisol concentration in lumbar
CSF are in broad agreement with those reported by
Murphy et al. (1967) for a similar variety of patients
suffering from various non-inflammatory neurological disorders, though Uete, Nishimura, Ohya,
Shimomura, and Tatebayashi (1970), who used a
fluorimetric method, have reported somewhat
lower control figures including those from healthy
subjects.
The failure to find any significant shift in the mean
I
6pm
8pm
CSF cortisol concentration, either upwards or
downwards, from the control mean, in depressed or
manic patients is of considerable interest. Patients
with depression are indeed suffering, and some disorder of the brain-pituitary-adrenal cortex control
system would be expected, if only as a secondary
phenomenon related to stress. Our data do not
support the implications of McClure's observations
that the brain in depression is lacking the level of
cortisol necessary for its proper functioning
(McClure and Cleghorn, 1969), nor do they reflect
the increase in total corticosteroids in the plasma
found in some depressives by many investigators.
Moreover, we found no correlation between severity
of depression and CSF cortisol concentration. As
previous studies (Brooksbank and Coppen, 1967;
Sachar, 1967) have shown that plasma and urinary
corticosteroids in depressives tend to fall during the
first few days after admission to hospital before any
marked clinical improvement has occurred, the relationship between CSF cortisol level and the interval
between admission and lumbar puncture was
examined in the present series: again no definite
trend was discernible.
It is probable that the concentration of cortisol
in CSF is directly related to that of the non proteinbound fraction in plasma (Murphy et al., 1967;
Uete et al., 1970). Unbound cortisol is likely to be
the form that is physiologically active in the brain
(Yates and Urquhart, 1962; Kawai and Yates, 1966):
the impermeability of the blood-brain barrier to
protein would seem to preclude any cerebral effect
Downloaded from http://jnnp.bmj.com/ on June 18, 2017 - Published by group.bmj.com
Cortisol in the cerebrospinal fluid of patients suffering from affective disorders
of protein-bound cortisol (Keller, Richardson, and
Yates, 1969). Our results therefore indicate that the
physiologically active level of cortisol in the brain
as a whole is unlikely to be grossly abnormal in
affective illness.
Uete and his collaborators (1970) did observe a
diurnal variation in the level of cortisol in the CSF
of five human subjects in whom continual sampling
of CSF was possible, but the fluctuations were small
compared with those found simultaneously in the
blood plasma. It is therefore not surprising that we
failed to find a definite relationship between time of
day and CSF cortisol concentration in our samples
taken from different individuals. The fact that the
CSF cortisol concentration in our control males,
who after all were not healthy, was apparently
correlated negatively with age should be noted with
caution. Total plasma cortisol and cortisol secretion rate are not lower in the elderly when allowance
is made for changes in body mass (Bliss, Sandberg,
Nelson, and Eik-Nes, 1953; Romanoff, Morris,
Welch, Rodriguez, and Pincus, 1961).
We wish to thank our colleagues at the Atkinson-Morley
and Graylingwell Hospitals for their collaboration in the
collection of samples of CSF, and Mr. J. Bailey for his
assistance with the statistical computations.
REFERENCES
Beardwell, C. G., Burke, C. W., and Cope, C. L. (1968).
Urinary free cortisol measured by competitive protein
binding. J. Endocr., 42, 79-89.
Beck, A. J., Ward, C. H., Mendelson, M., Mock, J., and
Erbaugh, J. (1961). An inventory for measuring depression.
Arch. gen. Psychiat. (Chic.), 4, 561-571.
Bliss, E. L., Sandberg, A. A., Nelson, D. H., and Eik-Nes, K.
(1953). The normal levels of 17-hydroxycorticosteroids in
the peripheral blood of man. J. clin. Invest., 32, 818-823.
Brooksbank, B. W. L., and Coppen, A. (1967). Plasma 11hydroxycorticosteroids in affective disorders. Brit. J.
Psychiat., 113, 395-404.
Coppen, A. J. (1968). Depressed states and indolealkylamines.
Adv'anc. Pharmacol., 6B, 283-291.
435
Coppen, A. (1970). The chemical pathology of the affective
disorders, pp. 189-210. Scientific Basis of Medicine Annual
Reviews.
Curzon, G. (1969). Tryptophan pyrrolase-a biochemical
factor in depressive illness? Brit. J. Psychiat., 115, 13671374.
Hamilton, M. (1967). Development of a rating scale for
primary depressive illness. Brit. J. clin. soc. Psychol., 6,
278-296.
Kawai, A., and Yates, F. E. (1966). Interference with feedback inhibition of adrenocorticotrophin release by protein
binding of corticosterone. Endocrinology, 79, 1040-1046.
Keller, N., Richardson, U. I., and Yates, F. E. (1969).
Protein binding and the biological activity of corticosteroids: in vivo induction of hepatic and pancreatic
alanine aminotransferases by corticosteroids in normal
and estrogen-treated rats. Endocrinology, 84, 49-62.
Lapin, I. P., and Oxenkrug, G. F. (1969). Intensification of
the central serotoninergic processes as a possible determinant of the thymoleptic effect. Lancet, 1, 132-136.
McClure, D. J. (1969). Hormonal Studies in the Cerebrospinal Fluid of Depressed Patients. Canadian Psychiatric
Association meeting, Toronto, June 13th.
McClure, D. J., and Cleghorn, R. A. (1969). Hormone
imbalance in depressive states. In The Future of the Brain
Sciences, pp. 525-553. Proceedings of a Conference.
Edited by S. Bogoch. Plenum Press: New York.
Murphy, B. E. P. (1967). Some studies of the protein-binding
of steroids and their application to the routine micro and
ultramicro measurement of various steroids in body fluids
by competitive protein-binding radioassay. J. clin. Endocr.,
27, 973-990.
Murphy, B. E. P., Cosgrove, J. B., Mcllquham, M. C., and
Pattee, C. J. Adrenal corticoid levels in human cerebrospinal fluid. Canad. med. Assoc. J., 97, 13-17.
Romanoff, L. P., Morris, C. W., Welch, P., Rodriguez, R. M.,
and Pincus, G. (1961). The metabolism of cortisol-4-C'4
in young and elderly men. I. Secretion rate of cortisol and
daily excretion of tetrahydrocortisol, allotetrahydrocortisol, tetrahydrocortisone and cortolone (20a and 20f).
J. clin. Endocr., 21, 1413-1425.
Sachar, E. J. (1967). Corticosteroids in depressive illness.
1. A re-evaluation of control issues and the literature. 11.
A longitudinal psychoendocrine study. Arch. gen. Psvchiat.,
17, 544-553, 554-567.
Uete, T., Nishimura, S., Ohya, H., Shimomura, T., and
Tatebayashi, Y. (1970). Corticosteroid levels in blood and
cerebrospinal fluid in various diseases. J. clin. Endocr.,
30, 208-214.
Yates, F. E., and Urquhart, J. (1962). Control of plasma
concentrations of adrenocortical hormones. Phisiol. Rev.,
42, 359-443.
Downloaded from http://jnnp.bmj.com/ on June 18, 2017 - Published by group.bmj.com
Cortisol in the cerebrospinal fluid
of patients suffering from
affective disorders
Alec Coppen, B. W. L. Brooksbank, R. Noguera and
D. A. Wilson
J Neurol Neurosurg Psychiatry 1971 34: 432-435
doi: 10.1136/jnnp.34.4.432
Updated information and services can be found at:
http://jnnp.bmj.com/content/34/4/432
These include:
Email alerting
service
Receive free email alerts when new articles cite this
article. Sign up in the box at the top right corner of the
online article.
Notes
To request permissions go to:
http://group.bmj.com/group/rights-licensing/permissions
To order reprints go to:
http://journals.bmj.com/cgi/reprintform
To subscribe to BMJ go to:
http://group.bmj.com/subscribe/