Changes in cortisol secretion during antidepressive
treatment and cognitive improvement in patients
with major depression: A longitudinal study
Kim Hinkelmann*, Steffen Moritz, Johannes Botzenhardt, Christoph Muhtz,
Klaus Wiedemann, Michael Kellner, Christian Otte
Department of Psychiatry and Psychotherapy, University Medical Center Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg,
Received 24 May 2011; received in revised form 27 August 2011; accepted 29 August 2011
Cognitive deficits are characteristic features of depression
and increased activity of the hypothalamus—pituitary—adre-
nal axis (HPA) leading to elevated cortisol is often reported in
major depression (Belanoff et al., 2001; de Kloet et al.,
2005a,b) although this depends on type and severity of
depression as well as on sample characteristics (Stetler
Psychoneuroendocrinology (2012) 37, 685—692
Objectives: We have previously reported that cognitive deficits are cross-sectionally associated
with elevated cortisol in depressed patients. Here, we longitudinally examined if changes in
cortisol secretion during treatment are associated with improvement of cognition.
Methods: Cognitive function and salivary cortisol levels were longitudinally examined in 52
patients with major depression before and after 3 weeks of standardized selective serotonin
reuptake inhibitor (SSRI) and an add-on treatment modulating the mineralocorticoid receptor and
compared to a healthy control group (n = 50) matched for age, gender and years of education.
Results: Across add-on treatment groups, SSRI treatment reduced salivary cortisol in patients to
levels of healthy controls (time ? group interaction p = .05). In patients, reduction of cortisol
significantly correlated with improvement in depressive symptoms (r = .52, p < .01), speed of
information processing (r = .50, p < .01), and cognitive set-shifting (r = .34, p = .03). Improved
depressive symptoms were only associated with improved attention and working memory.
Conclusions: Improvement of some cognitive domains during SSRI treatment was associated with
decreasing cortisol secretion and was only to a lesser extent associated with improved depressive
# 2011 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +49 40 7410 54222;
fax: +49 40 7410 53461.
E-mail address: email@example.com (K. Hinkelmann).
Available online at www.sciencedirect.com
j our na l h omepa g e: www.e lse vie r.c om/l oca te/ psyne ue n
0306-4530/$ — see front matter # 2011 Elsevier Ltd. All rights reserved.
and Miller, 2011; Knorr et al., 2010). We have recently shown
in a cross-sectional study that cognitive deficits are related to
elevated cortisol in depressed patients (Hinkelmann et al.,
2009) and several studies have shown a normalization of HPA
axis function during the course of treatment (McKay and
However, whether HPA normalization is related to
improvement of cognitive deficits, is still under debate.
To our knowledge, only two studies have longitudinally
examined the association between HPA normalization
and improvement of cognition in depressed patients com-
pared to healthy subjects: O‘Brien et al. (2004) found a
decrease of free cortisol in older depressed patients (mean
age 74 years) during treatment, which was not correlated
with cognitive improvement (O‘Brien et al., 2004). Com-
pared to healthy subjects, patients showed only slight
cognitive improvement and about half of them fulfilled
formal criteria for Mild Cognitive Impairment (MCI) during
Vythilingam et al. (2004) investigated a younger and
medication free sample of depressed patients in comparison
to healthy subjects. During SSRI treatment, patients showed
a decrease of cortisol and an improvement in memory func-
tion. However, at baseline, patients did not exhibit greater
cortisol levels compared to healthy controls and cortisol was
not associated with cognition before treatment. During fol-
low-up no healthy subjects were included and no association
between cognition and cortisol was found.
Two further studies longitudinally investigated the rela-
tionship of cortisol and cognition in depressed patients with-
out including a healthy control group: Zobel et al. (2004)
demonstrated an association of the decrease in cortisol and
the improvement of working memory function, which was
independent of symptom severity. In contrast, Reppermund
et al. (2007) did not find such associations but reported that a
high rate of depressed inpatients remained cognitively
impaired at discharge.
In summary, the few existing studies regarding HPA-axis
function and cognition during treatment in patients with
major depression have come to inconclusive results and
exhibit several limitations. In our previous cross-sectional
report from the same sample (Hinkelmann et al., 2009),
elevated cortisol was specifically associated with cognitive
deficits in verbal and non-verbal memory and in executive
function. Here, we aimed to further investigate the long-
itudinal association between changes of HPA-axis activity and
cognition in depressed patients during treatment. We
hypothesized that decreases in cortisol secretion during
antidepressant treatment would be associated with improve-
ment of cognitive function in the same domains, i.e. in verbal
and non-verbal memory and in executive function.
Clinical and demographic characteristics of the participants
are shown in Table 1. In brief, we recruited 52 in- and
outpatients (15 men and 37 women, mean age 35 ? 11.5
years, mean years of education 11.3 ? 1.6 years, Hamilton
Rating Scale for Depression score mean 27.2 ? 4.5, mean
number of episodes 1 ? 1.2) from a specialized depression
clinic at the Department of Psychiatry and Psychotherapy,
University Medical Center Hamburg. Inclusion criteria were
(1) a diagnosis of major depressive disorder, single or recur-
rent according to DSM-IV criteria, according to MINI-inter-
view and a minimum baseline score of 18 points on the
Patients (N = 52)
Controls (N = 50)
Non-psychotropic concomitant medication
Non-steroidal anti-inflammatory medication
Escitalopram dosage (mean mg/day 21)**
Lorazepam dosage (mean mg/day 21)
Zolpidem/Zopiclon dosage (mean no. tablets/day)
N = 6
N = 5
Abbreviations: BMI: Body Mass Index, BDI: Beck Depression Inventory.
*Based on independent t-test for continuous variables and chi-square for dichotomous variables.
**At baseline all patients were free of psychotropic medication and were all started on escitalopram after the baseline assessment.
K. Hinkelmann et al.
Hamilton Rating Scale for Depression, 17-item version
(HAMD-17); (2) age from 18 to 65 years; (3) a period of at
least 5 days free from antidepressants, antipsychotics, mood
stabilizers, and other medications influencing HPA activity.
43% of the patients referred were first episode patients. All
patients were referred either untreated or with major
depression despite medication. Only 7 patients received
psychotropic treatment (except benzodiazepines) when
referred to the clinic and needed to go through a washed
out: three patients took mirtazapine, 1 trazodone, 1 parox-
etine, 1 escitalopram and 1 buspirone.
Criteria for exclusion were (1) dementia, schizophrenia
spectrum disorder, bipolar disorder, substance dependence
within the last 6 months according to the MINI-interview
(Sheehan et al., 1998), (2) serious medical conditions, espe-
cially those associated with adrenal dysfunctions; steroid use
or well known impact on HPA activity (e.g. diabetes mellitus)
or cognitive function, (3) pregnancy and nursing, and (4)
fluoxetine medication due to long half life time.
A control group of 50 healthy subjects (15 males and 35
females) recruited by public postings and matched for age (?
3 years), sex and years of education (1 to 1 matching) were
enrolled in the study. Subjects were free of former and
present DSM-IV axis I disorders according to the MINI-inter-
view, had no physical illness and had been free of any
medication at least 3 months.
The study was approved by the local ethics committee.
After complete description of the study to the subjects,
written informed consent was obtained.
Cortisol data were available for all 52 patients and 50
controls at both time points. Out of 102 participants, who all
completed day 1, 47 healthy controls and 44 patients com-
pleted both days of the neuropsychological assessment.
2.2. Design and procedure
All participants were tested twice: at baseline (day 0) and
after 3 weeks. Prior to study enrollment, there was a 5-days
wash-out period for those patients who received antidepres-
sive medication and for whom we decided to switch anti-
depressive medication due to clinical reasons (no patient was
tapered off from successful medication to be enrolled into
the study) to avoid the effects of antidepressant discontinua-
tion on cortisol secretion (Michelson et al., 2000). After the
wash-out period, all patients were tested without antide-
pressant during the baseline testing. After the baseline
assessment, all patients started treatment with escitalopram
10 mg that could be increased to 15 or 20 mg during the
following weeks. Additionally, as this study was part of a
randomized, double-blind controlled trial with a 2:2:1 ran-
domization, each patient received an add-on treatment for 3
weeks modulating the mineralocorticoid receptor (19
patients received the MR-agonist fludrocortisone, 22 the
MR-antagonist spironolactone and 11 received placebo as
add-on treatment) as described in (Otte et al., 2010).
2.3. Hormonal assessment
Salivary cortisol was collected at 0800 h, 1200 h, 1600 h and
2200 h on day 0 and day 21. All participants received oral and
written instructions on the correct use of the Salivette
salivary collection device (Sarstedt, Germany). Participants
were advised not to eat, drink, smoke, brush their teeth, or
use mouthwash in the 30 min prior to salivary collection.
Cortisol was determined by radioimmunoassay (DRG, Mar-
burg, Germany). Inter- and intraassays coefficients of varia-
tion were below 8%. Detection limits were 0.5 ng/ml for
None of the cortisol values were below the detection
2.4. Neuropsychological assessment
Neuropsychological tests were conducted at baseline and day
21 and comprised the Auditory Verbal Learning Test (AVLT),
the Digit Span Test, Rey/Taylor Figures, letter cancellation
test (d2) and the Trail Making Test A and B (TMT-A, TMT-B).
All neuropsychological tests were administered by trained
psychologists. The time of testing was held constant between
1400 h and 1600 h.
Auditory verbal learning test (AVLT) (Lezak, 1995): the
AVLT is a measure of short-term and long-term verbal
memory. The experimenter reads a list of 15 words (list
A), which the participant is requested to repeat in loose
order. After list A has been presented five times, the subject
is asked to reproduce words from a newly presented list (list
B). Following this, the subject is instructed to recall the
words from list A without renewed presentation. After
30 min, the subject is again asked to repeat the words from
list A. Outcome measure is the number of correctly remem-
bered words after each presentation, the delayed recall
after 30 min for long term memory and the sum of all
presentations (total score).
Trail-making test (TMT) (Reitan, 1992): speed of cognitive
processing was assessed with the TMT part A. In this task, the
subject has to connect encircled numbers in ascending order
as quickly as possible. Part B assesses executive function and
requires the alternation between numbers and letters in
ascending order. The score of each part is represented by
the time needed to complete the task. In addition to direct
scores, the difference score (B—A) reflects switch cost, a
relatively pure indicator of executive function (Sanchez-
Cubillo et al., 2009).
Forward and backward digit span (Tewes, 1991): this task
forms part of the Wechsler Adult Intelligence Scale (WAIS).
During the forward digit span task, participants are asked to
remember a series of digits and repeat them back in the same
order. During the backward digit span task, they are asked to
repeat the digits in reverse order, which taps working mem-
ory. Outcome measure is the number of correctly remem-
Rey—Osterrieth complex figure test (RCFT) and Taylor
complex figure test (TCFT) (Osterrieth, 1944): these tests
measure visuospatial memory. The participant is first
required to copy a complex figure. Immediately thereafter
and 20 min later the figure has to be re-drawn from memory.
Outcome measure is the number of correctly remembered
items in each condition (copy, immediate recall and delayed
recall). To adjust for baseline performance and extract a true
memory parameter, we adjusted the immediate and delayed
recall by the performance on the first copy trial (‘‘baseline
Cortisol and cognition in the course of MDD
Test d2 (Brickenkamp, 1978): The Test d2 is a letter
cancellation test that taps selective attention/concentra-
tion. In this task, the subject is instructed to cross out the
letter d whenever it is accompanied by two small lines; d’s
with more than or less than two lines or any stimuli containing
the character p serve as distracters. Subsequent to a practice
trial, 14 rows with target and distracter stimuli are pre-
sented. The attention score is the number of correctly
2.5. Statistical analyses
The cortisol and cognitive measures were normally distrib-
uted. Mixed analyses of covariance (ANCOVA) with repeated
measures design were conducted to investigate differences
in cortisol levels and neuropsychological tests with group
(patients vs. healthy subjects) as between-subjects factor
and time (day 0 vs. day 21) as within-subjects factor adjust-
ing for potentially confounding variables. For cortisol secre-
tion during the day, we also calculated the area under the
curve with respect to ground (AUC) as described in (Pruessner
et al., 2003). Differences between patients and controls in
cortisol AUC for day 0 and day 21 were then analyzed in rm-
To investigate the effect of the add-on treatment on
cognitive measures within the patient group mixed analyses
of variance (ANCOVA), adjusted for age and education, with
repeated measures design were conducted with treatment
(spironolactone vs. fludrocortisone vs. placebo) as between-
subjects factor and time (day 0 vs. day 21) as within-subjects
Pearson correlation coefficients were determined in the
total sample and in each group separately to examine the
association between the change of cortisol secretion and
change in cognitive function. In order to control for poten-
tially confounding variables, we also conducted linear regres-
sion analyses for neuropsychological tests. A sample size of
n = 52 is sufficient to detect the effect size of f2 = 0.25
(medium effect size) with 80% power in linear regression
analyses with four predictors.
We adjusted all analyses comparing patients with controls
for smoking because depressed patients were more likely to
smoke as described previously (Hinkelmann et al., 2009).
In all analyses, two-sided tests were used and as nominal
level of significance, a = 0.05 was accepted while a ? 0.10
was considered a trend.
3.1. Cortisol secretion
Three weeks of antidepressant treatment reduced salivary
cortisol in patients to levels of age- and sex matched healthy
subjects (repeated-measures ANCOVA, adjusted for smoking,
time ? group interaction F(1;99) = 3.8, p = .05, partial Eta
squared = .04) (see Fig. 1).
3.2. Cognitive function
As indicated in Table 2, both patients and healthy controls
improved in most cognitive variables over time. However,
patients improved to a greater extent in verbal memory
(repeated-measures ANCOVA, adjusted for smoking, group -
? time interaction, AVLT long term memory: F(1;90) = 5.1,
p = .02, partial Eta squared = .06) and non-verbal memory
(repeated-measures ANOVA group ? time interaction, Rey/
Taylor Figure: F(1;90) = 8.8, p < .01, partial Eta squared =
.09) compared to healthy controls.
In fact, after three weeks patients still performed worse
compared to healthy controls (ANCOVA, adjusted for smok-
ing, significant for Digit Span Forward: F(1;90) = 5.0, p = .02,
squared = .05;
p < .01, partial Eta squared = .23 and d2: F(1;90) = 8.9,
p < .01, partial Eta squared = .09) in comparison to healthy
subjects (see Table 2) indicating ongoing relative cognitive
deficits in patients compared to controls in these domains.
Repeated-measures ANCOVA (adjusted for age and educa-
tion) revealed no differences between the three treatment
groups over time (repeated-measures ANCOVA, p > .05 for all
time (baseline vs. day 21) ? treatment (fludrocortisone vs.
spironolactone vs. placebo) interactions).
F(1;90) = 26.1,
3.3. Cortisol secretion and cognitive function
In patients, reduction of cortisol was significantly associated
with improved cognitive function in TMT-A (r = .50, p < .01)
and TMT-B (r = .34, p = .03). Noteworthy, reduction of corti-
sol showed a trend towards an association with AVLT total
score (r = .26, p = .09), TMT-Difference Score (B—A) (r = .25,
p = .10), Rey/Taylor Figure copy (r = ?.28, p = .07), and Digit
Span Backward (r = .27, p = .08) (see Table 3). As indicated in
Table 3, the correlations in the total sample were driven by
the depressed patients.
The add-on treatment had no effect on the cognitive
measures. However, to control for any residual confounding
and to further control for other potentially confounding
(n = 52) before and after a 3-weeks treatment with escitalopram
compared to healthy subjects (n = 50). Bars represent mean area
under the curve (AUC) cortisol values. Repeated-measures
( p = .05). Post hoc tests revealed a significant reduction of
cortisol in patients (t-test, t = 2.5, df = 51, p = 0.01).
Salivary cortisol (area under the curve) of patients
time ? group
K. Hinkelmann et al.
variables, linear regression analyses for each cognitive test as
outcome variable and with age, sex and add-on treatment as
predictor variables were conducted for the patient group.
Adjusting for age, sex and add-on treatment, delta AUC
Cortisol was significantly associated with TMT-A and Rey/
Taylor copy ( p < .05) and on trend level with AVLT total
score, TMT-B, d2, Rey/Taylor immediate and delayed recall
(all p-values < 10).
Three weeks of antidepressant treatment significantly
improved depression (HAMD day 0 = mean 27.2 ? SD 4.5 vs.
day 21 15.2 ? SD 7.3, paired t-test, p < .01, effect size
r = .71). Importantly, improvement of psychopathology was
strongly correlated with decreases in cortisol (r = 52, p < 01)
(see Fig. 2), but only moderately correlated with cognitive
improvement (significant only for the letter cancellation test
(r = ?.35, p = .02) and digit span backward (r = .33, p = .03).
In this study, we longitudinally examined if reduction of
cortisol is associated with improvement of cognitive function
in the course of treatment of depressed patients. We found
that antidepressant treatment reduced salivary cortisol in
depressed patients to the level of healthy subjects within
three weeks. The reduction of cortisol was related to
improvement of some but not all cognitive domains. Sec-
ondly, the reduction of cortisol was correlated with improve-
ment of psychopathology. However, cognitive improvement
was not accounted for by psychopathological improvement.
Correlation analysis of change scores in salivary cortisol and neuropsychological tests.
Both groups (n = 91)
D AUC cortisol r, ( p)
Patients (n = 44)
D AUC cortisol r, ( p)
Healthy subjects (n = 47)
D AUC cortisol r, ( p)
Cognitive domain (test)
D AVLT total Score
D AVLT delayed recall
D TMT-diff (B—A)
D Digit span forward
D Digit span backward
D Rey/Taylor copy
D Rey/Taylor immediate recall (baseline corrected)
D Rey/Taylor delayed recall (baseline corrected)
D d2_concentration score
0.50 (< 0.01)*
AUC = area under the curve, p = probability.
*p < 0.05.
**p < 0.1.
Mean scores (SE) of cognitive variables for patients before and after treatment compared to healthy subjects.
(n = 44)
(n = 47)
Digit span forward
Digit span backward
Rey/Taylor immediate recall
Rey/Taylor delayed recall
d2 Concentration score
Patients improved to a greater extent in verbal memory (repeated-measures ANCOVA group ? time interaction, AVLT long term memory:
F(1;90) = 5.1, p = .02) and non-verbal memory (repeated-measures ANCOVA group ? time interaction, Rey/Taylor Figure: F(1;90) = 8.8,
p < .01) compared to healthy controls.
AVLT: Auditory Verbal Learning Test, TMT A and B: Trail Making Test A and B.
Cortisol and cognition in the course of MDD
It has long been hypothesized that increased cortisol
secretion might be responsible for cognitive deficits in
depressed patients (Belanoff et al., 2001; de Kloet et al.,
2005a). However, to the best of our knowledge, this is the
first study demonstrating the association of HPA normaliza-
tion with cognitive improvement in some but not all domains
in depressed patients compared to healthy controls. Our
results suggest that lowering cortisol or blocking its effects
might be a promising approach in the future treatment of
patients with major depression.
Our results are in line with Zobel et al. (2004) who found
cortisol reduction during antidepressant treatment to be
correlated with improvement in cognitive function in
patients with major depression. However, whereas Zobel
et al. found an association with working memory, in our study
cortisol reduction was significantly associated only with
executive function, while verbal and non-verbal memory
were suggestive of an association, but did not reach statis-
The few existing other prospective studies have come to
results challenging the theory of a causal role of cortisol
regarding cognitive deficits (Reppermund et al., 2007; O‘Brien
et al., 2004; Vythilingam et al., 2004) but these studies
differed in some aspects from our study. In some studies,
patients were already on medication during the baseline
assessment (Reppermund et al., 2007; O‘Brien et al., 2004).
Other studies did not include a healthy control group (Repper-
mund et al., 2007), did focus on elderly patients only that were
to great extent cognitively impaired independent from depres-
sion (O‘Brien et al., 2004), or patients did not show elevated
cortisol compared to healthy controls (Vythilingam et al.,
2004). Thus, future studies should examine to what extent
medication, in- and exclusion criteria or gender aspects mod-
erate the longitudinal association between cortisol secretion
and cognitive function in depressed patients.
Our results suggest that the improvement of cognitive
function after restoring the HPA axis is not specific for a single
cognitive function since more cognitive domains improved.
The strongest correlation with cortisol reduction in our
sample was with speed of information processing (TMT-A)
and executive function (TMT-B), whereas memory functions
were only related on trend level. The TMT-A, a test demand-
ing visual search abilities (Sanchez-Cubillo et al., 2009), has
been linked to frontal lobe impairment (Demakis, 2004), but
also exhibits intercorrelations with working memory func-
tions (Sanchez-Cubillo et al., 2009).
In our study, the decrease of cortisol was correlated with
improvement in cognition and improvement in psychopathol-
ogy. The latter two, however, were only moderately asso-
ciated with each other. This suggests that improvement of
cognition is not just a function of improvement in psycho-
pathology which is in line with Zobel et al. (2004) but in
contrast to Reppermund et al. (2007), who found the
improvement of psychopathology to be related with speed
of information processing. Although our results indicate that
cortisol secretion is associated with changes in cognitive
function above and beyond improved psychopathology,
future longitudinal studies with repeated assessments and
a more extended observation period are needed to disen-
tangle the temporal associations among changes in cortisol,
cognition, and psychopathology during treatment.
In our study, participants improved across groups in cog-
nitive variables indicating a learning effect. We therefore
conclude that it will be important to include healthy control
groups in order to control for this effect in future studies.
Interestingly, compared to healthy subjects patients
improved significantly stronger in verbal and non-verbal
memory functions, which are hippocampus related domains.
These results are consistent with the hypothesis that ele-
vated cortisol exerts detrimental effects on the hippocampus
(de Kloet et al., 2005b), the brain region of highest gluco-
corticoid and mineralocorticoid receptor density. However, it
should be kept in mind that after three weeks of treatment
patients still fared worse in the Digit Span forward test
(working memory) and the Rey/Taylor Figure (non-verbal
memory), which is in line with Douglas et al. (2011).
Our results are also consistent with other data supporting
a fundamental role of elevated cortisol in cognitive deficits:
acute and chronic glucocorticoid treatment leads to cogni-
tive impairments in healthy subjects (Kirschbaum et al.,
1996; Lupien and McEwen, 1997; Young et al., 1999; New-
comer et al., 1999) as well as to impaired cognitive function
in patients on long-term prescription corticosteroid therapy
(Brown et al., 2004). Patients with glucocorticoid excess due
to Cushing’s disease exhibit worse cognitive function (Forget
et al., 2000; Starkman et al., 2001) and improve after
successful treatment (Starkman et al., 2003; Hook et al.,
2007). And lastly, spatial working memory, verbal fluency,
and spatial recognition memory improved in depressed bipo-
lar patients after blocking the glucocorticoid receptor (Young
et al., 2004).
Our study had several strengths. We examined ‘‘real-
world’’, treatment-seeking patients with moderate to severe
depression and all patients were studied at baseline not on
psychotropic medication for at least five days. However,
several limitations must also be acknowledged: our correla-
tional statistical methods do not allow definite inferences
about causality. Only two cognitive variables were signifi-
cantly correlated with the decrease of cortisol, whereas five
cognitive variables were suggestive of an association, but did
not reach statistical significance. Therefore, we were only
partially able to confirm our hypothesis. However, p-values
Hamilton Depression Rating Scale in patients. Pearson correla-
tion: r = 0.52, p < 0.01, AUC = area under the curve for cortisol.
Correlation between changes in salivary cortisol and
K. Hinkelmann et al.
depend on effects sizes and sample size. For example, in our
study, an r-value of 0.25 is significant in the total sample,
whereas the same r-value is a weak trend in the depressed
group due to smaller sample size. For the delta of Rey/Taylor
copy, the r-value is 0.24 in the total group and significant but
0.28 in the depressed group and non-significant. There are
more examples like these in our study (see Table 3). We think
it would be unfortunate to interpret an r-value of 0.24 in the
total group (because it is significant) but not to interpret an r-
value of 0.28 (because it is non-significant). However, due to
multiple comparisons the conclusions that can be drawn are
Secondly, patients were treated with an add-on medica-
tion modulating the mineralocorticoid receptor during the
first three weeks of treatment as described in Otte et al.
(2010), which might have an impact on cognitive function.
However, this was not the case in this study. We nevertheless
considered the add-on treatment as a covariate in regression
analyses to avoid any residual confounding. Furthermore, if
anything, the ‘‘noise’’ we hereby induced in the association
between changes in cortisol secretion and improvement in
cognitive function HPA-axis function would have biased our
results towards the null hypothesis.
Furthermore, alterations in executive function are com-
mon in depressed patients and we used only one measure
(TMT-B) to specifically examine executive function.
Three weeks of follow-up are a relatively short period of
time to assess changes in cognitive function in depressed
patients some of whom were still not in remission. However,
in this time span salivary cortisol was significantly reduced to
the level of healthy subjects and effects in cognition could be
seen, as in other studies with short observation periods (Zobel
et al., 2004). Nevertheless, this study was not designed to
assess the question if cognitive deficits are state or trait
related in MDD. Therefore, further studies should systemati-
cally explore the effects of treatment-resistance, response,
and remission on changes in cognitive function in depressed
patients over a longer period of time. Finally, we did not
investigate the cortisol awakening response (Fries et al.,
2009). Differences in cortisol level between patients and
healthy subjects as well as correlations might have been even
more pronounced using this measure (Vreeburg et al., 2009).
In summary, our results demonstrate that the normal-
ization of the HPA axis activity is associated with the improve-
ment of cognition in patients with major depression during
antidepressant treatment. HPA activity might therefore be a
promising target to treat cognitive dysfunction in major
KH and CO designed the study, wrote the protocol and the
manuscript. JB, KH and CM examined the subjects and
collected the data. MK, KW and SM helped to manage the
analyses and to interpret the data.
Role of funding source
This work was supported by the German Research Foundation
(Deutsche Forschungsgemeinschaft, grant OT 209/3-1, 3-2).
The German Research Foundation had no role in the collection
of data, interpretation of results, or preparation of this manu-
script. We have no conflict of interest.
Conflict of interest
Dr. Wiedemann served as a consultant to, or has been on the
speakers boards of AstraZeneca, BristolMyersSquibb, Jans-
sen, Pzer, Servier and Wyeth. Dr. Kellner received funding for
investigator initiated trials by Lundbeck and Pzer. He is a
member of an advisory board for Wyeth. Dr. Otte is on the
speaker’s board of Astra Zeneca, Lundbeck and Servier. Dr.
Hinkelmann, Dr. Moritz, Dr. Muhtz and Dr. Botzenhardt report
no conflict of interest.
We are grateful to the excellent technical assistance of Iris
Remmlinger-Marten and Kirsten Huwald.
Belanoff, J.K., Gross, K., Yager, A., Schatzberg, A.F., 2001. Cortico-
steroids and cognition. J. Psychiatr. Res. 35, 127—145.
Brickenkamp, R., 1978. Test d2 Handanweisung (Test d2. Manual)
Hogrefe, Go ¨ttingen, Germany.
Brown, E.S., D, J.W., Frol, A., Bobadilla, L., Khan, D.A., Hanczyc, M.,
Rush, A.J., Fleckenstein, J., Babcock, E., Cullum, C.M., 2004.
Hippocampal volume, spectroscopy, cognition, and mood in
patients receiving corticosteroid therapy. Biol. Psychiatry 55,
de Kloet, E.R., Joels, M., Holsboer, F., 2005a. Stress and the brain:
from adaptation to disease. Nat. Rev. Neurosci. 6, 463—475.
de Kloet, E.R., Sibug, R.M., Helmerhorst, F.M., Schmidt, M.V., 2005b.
Stress, genes and the mechanism of programming the brain for
later life. Neurosci. Biobehav. Rev. 29, 271—281.
Demakis, G.J., 2004. Frontal lobe damage and tests of executive
processing: a meta-analysis of the category test, stroop test, and
trail-making test. J. Clin. Exp. Neuropsychol. 26, 441—450.
Douglas, K.M., Porter, R.J., Knight, R.G., Maruff, P., 2011. Neuro-
psychological changes and treatment response in severe depres-
sion. Br. J. Psychiatry 198, 115—122.
Forget, H., Lacroix, A., Somma, M., Cohen, H., 2000. Cognitive
decline in patients with Cushing’s syndrome. J. Int. Neuropsychol.
Soc. 6, 20—29.
Fries, E., Dettenborn, L., Kirschbaum, C., 2009. The cortisol awak-
ening response (CAR): facts and future directions. Int. J. Psycho-
physiol. 72, 67—73.
Hinkelmann, K., Moritz, S., Botzenhardt, J., Riedesel, K., Wiede-
mann, K., Kellner, M., Otte, C., 2009. Cognitive impairment in
major depression: association with salivary cortisol. Biol. Psychi-
atry 66, 879—885.
Hook, J.N., Giordani, B., Schteingart, D.E., Guire, K., Giles, J., Ryan,
K., Gebarski, S.S., Langenecker, S.A., Starkman, M.N., 2007.
Patterns of cognitive change over time and relationship to age
following successful treatment of Cushing’s disease. J. Int. Neu-
ropsychol. Soc. 13, 21—29.
Kirschbaum, C., Wolf, O.T., May, M., Wippich, W., Hellhammer, D.H.,
1996. Stress- and treatment-induced elevations of cortisol levels
associated with impaired declarative memory in healthy adults.
Life Sci. 58, 1475—1483.
Knorr, U., Vinberg, M., Kessing, L.V., Wetterslev, J., 2010. Salivary
cortisol in depressed patients versus control persons: a system-
atic review and meta-analysis. Psychoneuroendocrinology 35,
Cortisol and cognition in the course of MDD
Lezak, M., 1995. Neuropsychological Assessment. Oxford University Download full-text
Press, New York.
Lupien, S.J., McEwen, B.S., 1997. The acute effects of corticoste-
roids on cognition: integration of animal and human model
studies. Brain Res. Rev. 24, 1—27.
McKay, M.S., Zakzanis, K.K., 2010. The impact of treatment on HPA
axis activity in unipolar major depression. J. Psychiatr. Res. 44,
Michelson, D., Amsterdam, J., Apter, J., Fava, M., Londborg, P.,
Tamura, R., Pagh, L., 2000. Hormonal markers of stress response
following interruption of selective serotonin reuptake inhibitor
treatment. Psychoneuroendocrinology 25, 169—177.
Newcomer, J.W., Selke, G., Melson, A.K., Hershey, T., Craft, S.,
Richards, K., Alderson, A.L., 1999. Decreased memory perfor-
mance in healthy humans induced by stress-level cortisol treat-
ment. Arch. Gen. Psychiatry 56, 527—533.
O‘Brien, J.T., Lloyd, A., McKeith, I., Gholkar, A., Ferrier, N., 2004. A
longitudinal study of hippocampal volume, cortisol levels, and
cognition in older depressed subjects. Am. J. Psychiatry 161,
Osterrieth, P.A., 1944. Le test de copie d’une figure complexe;con-
tribution a l’etude de la perception et de la memoire (test of
copying a complex figure; contribution to the study of perception
and memory). Arch. Psychol. 30, 206—356.
Otte, C., Hinkelmann, K., Yassouridis, A., Jahn, H., Wiedemann, K.,
Kellner, M., 2010. Stimulation vs. blockade of the mineralocorti-
coid receptor as adjunctive therapy to SSRI treatment of major
depression: randomized, double-blind, placebo-controlled trial.
J. Psychiatr. Res. 44, 339—346.
Pruessner, J.C., Kirschbaum, C., Meinlschmid, G., Hellhammer, D.H.,
2003. Two formulas for computation of the area under the curve
represent measures of total hormone concentration versus time-
dependent change. Psychoneuroendocrinology 28, 916—931.
Reitan, R.M., 1992. Trail Making Test Manual of Administration and
Scoring. Reitan Neuropsychology Laboratory, Tuscon, Arizona.
Reppermund, S., Zihl, J., Lucae, S., Horstmann, S., Kloiber, S.,
Holsboer, F., Ising, M., 2007. Persistent cognitive impairment in
depression: the role of psychopathology and altered hypothalam-
ic—pituitary—adrenocortical (HPA) system regulation. Biol. Psy-
chiatry 62, 400—406.
Sanchez-Cubillo, I., Perianez, J.A., Adrover-Roig, D., Rodriguez-
Sanchez, J.M., Rios-Lago, M., Tirapu, J., Barcelo, F., 2009.
Construct validity of the trail making test: role of task-switching,
working memory, inhibition/interference control, and visuomo-
tor abilities. J. Int. Neuropsychol. Soc. 15, 438—450.
Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J.,
Weiller, E., Hergueta, T., Baker, R., Dunbar, G.C., 1998. The
Mini-International Neuropsychiatric Interview (M.I.N.I.): the
development and validation of a structured diagnostic psychi-
atric interview for DSM-IV and ICD-10. J. Clin. Psychiatry 59
(Suppl. 20), 22—33 quiz 34—57.
Starkman, M.N., Giordani, B., Berent, S., Schork, M.A., Schteingart,
D.E., 2001. Elevated cortisol levels in Cushing’s disease are asso-
ciated with cognitive decrements. Psychosom. Med. 63, 985—993.
Starkman, M.N., Giordani, B., Gebarski, S.S., Schteingart, D.E., 2003.
Improvement in learning associated with increase in hippocampal
formation volume. Biol. Psychiatry 53,
Stetler, C., Miller, G.E., 2011. Depression and hypothalamic—pitui-
tary—adrenal activation: a quantitative summary of four decades
of research. Psychosom. Med. 73, 114—126.
Tewes, U., 1991. HAWIE-R: Hamburg-Wechsler Intelligenztest fu ¨r
Erwachsene; Handbuch und Testanweisung. Huber-Verlag, Bern.
Vreeburg, S.A., Hoogendijk, W.J., van Pelt, J., Derijk, R.H., Verha-
gen, J.C., van Dyck, R., Smit, J.H., Zitman, F.G., Penninx, B.W.,
2009. Major depressive disorder and hypothalamic—pituitary—
adrenal axis activity: results from a large cohort study. Arch.
Gen. Psychiatry 66, 617—626.
Vythilingam, M., Vermetten, E., Anderson, G.M., Luckenbaugh, D.,
Anderson, E.R., Snow, J., Staib, L.H., Charney, D.S., Bremner,
J.D., 2004. Hippocampal volume, memory, and cortisol status in
major depressive disorder: effects of treatment. Biol. Psychiatry
Young, A.H., Gallagher, P., Watson, S., Del-Estal, D., Owen, B.M.,
Ferrier, I.N., 2004. Improvements in neurocognitive function
and mood following adjunctive treatment with mifepristone
(RU-486) in bipolar disorder. Neuropsychopharmacology 29,
Young, A.H., Sahakian, B.J., Robbins, T.W., Cowen, P.J., 1999. The
effects of chronic administration of hydrocortisone on cognitive
function in normal male volunteers. Psychopharmacology (Berl)
Zobel, A.W., Schulze-Rauschenbach, S., von Widdern, O.C., Metten,
M., Freymann, N., Grasmader, K., Pfeiffer, U., Schnell, S.,
Wagner, M., Maier, W., 2004. Improvement of working but not
declarative memory is correlated with HPA normalization during
antidepressant treatment. J. Psychiatr. Res. 38, 377—383.
K. Hinkelmann et al.