brain research 1598 (2015) 1–11
Available online at www.sciencedirect.com
www.elsevier.com/locate/brainres
Research Report
The association between cortisol and the BOLD
response in male adolescents undergoing fMRI
Esther H.H. Keulersa,b,n, Peter Stiersa, Nancy A. Nicolsonb, Jelle Jollesc
a
Department of Neuropsychology and Psychopharmacology, Faculty of Psychology and Neuroscience, Maastricht
University, Maastricht, The Netherlands
b
Department of Psychiatry & Neuropsychology, School for Mental Health and Neuroscience, Faculty of Health,
Medicine and Life Sciences, Maastricht University, Maastricht, The Netherlands
c
Centre Brain and Learning and AZIRE Research Institute, Faculty of Psychology and Education, VU Amsterdam,
Amsterdam, The Netherlands
art i cle i nfo
ab st rac t
Article history:
MRI participation has been shown to induce subjective and neuroendocrine stress
Accepted 6 December 2014
reactions. A recent aging study showed that cortisol levels during fMRI have an age-
Available online 13 December 2014
dependent effect on cognitive performance and brain functioning. The present study
Keywords:
examined whether this age-specific influence of cortisol on behavioral and brain activation
Functional magnetic resonance
levels also applies to adolescence. Salivary cortisol as well as subjective experienced
imaging
anxiety were assessed during the practice session, at home, and before, during and after
Salivary cortisol
the fMRI session in young versus old male adolescents. Cortisol levels were enhanced pre-
HPA axis
imaging relative to during and post-imaging in both age groups, suggesting anticipatory
Adolescence
stress and anxiety. Overall, a negative correlation was found between cortisol output
Psychological stress and anxiety
during the fMRI experiment and brain activation magnitude during performance of a
Individual differences
gambling task. In young but not in old adolescents, higher cortisol output was related to
stronger deactivation of clusters in the anterior and posterior cingulate cortex. In old but
not in young adolescents, a negative correlation was found between cortisol and activation
in the inferior parietal and in the superior frontal cortex. In sum, cortisol increased the
deactivation of several brain areas, although the location of the affected areas in the brain
was age-dependent. The present findings suggest that cortisol output during fMRI should
be considered as confounder and integrated in analyzing developmental changes in brain
activation during adolescence.
& 2014 Elsevier B.V. All rights reserved.
n
Corresponding author at: Department of Neuropsychology and Psychopharmacology, Faculty of Psychology and Neuroscience,
Maastricht University, Universiteitssingel 40, PO Box 616, 6200 MD Maastricht, The Netherlands, Fax: þ31 43 3884560.
E-mail address: [email protected] (E.H.H. Keulers).
http://dx.doi.org/10.1016/j.brainres.2014.12.022
0006-8993/& 2014 Elsevier B.V. All rights reserved.
2
1.
brain research 1598 (2015) 1–11
Introduction
Participation in a functional magnetic resonance imaging
(fMRI) experiment has been shown to induce both subjective
and neuroendocrine stress reactions in adults (Melendez and
McCrank, 1993; Muehlhan et al., 2011). Self-reported stress
and anxiety are higher before than after the fMRI session
(Dantendorfer et al., 1997; Muehlhan et al., 2011). Interpreting
a situation as being stressful triggers the activation of the
hypothalamic–pituitary–adrenal (HPA) axis, which ultimately
results in the secretion of cortisol. Activation of the HPA axis
has been validated for, among others, motivated and challenging performance tasks, novelty and uncontrollability of a
situation, which is generally the case in an fMRI experiment
(Dickerson and Kemeny, 2004). Indeed, elevated levels of
cortisol have been associated with fMRI participation
(Muehlhan et al., 2011; Tessner et al., 2006).
Cortisol has been shown to influence cognition and brain
functioning. Cortisol can easily cross the blood–brain barrier
and binds to receptors in the brain, with a preference for
limbic system structures, including hippocampus and amygdala, and the prefrontal cortex (Lupien et al., 2007). Glucocorticoids have been shown to influence cognition according
to an inverted U-shaped curve (De Kloet et al., 1999).
Decreased as well as increased levels of cortisol due to either
pharmacological manipulations or laboratory stressors can
impair cognitive functions. For example, low levels of cortisol
after metyrapore treatment (inhibitor of cortisol synthesis)
impaired declarative memory compared with a placebo
treatment (Lupien et al., 2002; Maheu et al., 2005). Moreover,
elevated levels of cortisol induced by either stress tasks
(Buckert et al., 2014; Kirschbaum et al., 1996) or cortisone
treatment (De Quervain et al., 2000) impaired performance on
cognitive tasks measuring memory, spatial thinking and risk
taking. Several neuroimaging studies have found decreased
brain activation in medial and orbital prefrontal cortices
during stress paradigms (Kern et al., 2008; Pruessner et al.,
2008; Ossewaarde et al., 2011). Additionally, increased cortisol
levels have been associated with decreased activation in
limbic structures, such as the hippocampus and amygdala
(Dedovic et al., 2009; Pruessner et al., 2008). Positive associations have been established between cortisol levels and
activation in parietal, ventrolateral and dorsolateral prefrontal areas (Kern et al., 2008; Weerda et al., 2010).
Recently, some studies have shown that the neuroendocrine stress response associated with the fMRI participation
itself influences brain functioning (Kukolja et al., 2008;
Muehlhan et al., 2011). The experienced stress during an fMRI
experiment can influence the quality of the images, task
performance and functional activation patterns (Dantendorfer
et al., 1997; Eatough et al., 2009; Kukolja et al., 2008; Melendez
and McCrank, 1993). These unintended effects of the scanner
environment itself can act as a potential confounder in fMRI
research. Moreover, Kukolja et al. (2008) showed that the
cortisol response during an fMRI experiment has a differential
effect on behavior and brain functioning in young versus old
adults. Namely, cortisol response had a positive association
with accuracy and prefrontal activation during encoding in
young but not in old adults, whereas cortisol response was
negatively correlated with prefrontal cortex and hippocampus
activation during retrieval in old but not in young adults. This
suggests that considering cortisol as confounder might be
especially relevant in studies comparing different age groups.
In recent years, fMRI has increasingly been used to study
developmental changes, by comparing brain activation
between groups of children, adolescents and adults. We
hypothesize that developmental fMRI results might be confounded by cortisol output associated with fMRI participation,
especially in adolescent samples. Adolescence is associated
with an increased number of stressful life events as well as
an increased perception of occurrences as stressful relative to
child and adulthood (Larson and Asmussen, 1991). Additionally, older adolescents, aged 15–17 years, have shown an
increased cortisol responsivity to psychological stressors as
well as increased basal cortisol levels compared with children
and young adolescents aged 7–13 years (Adam, 2006; Gunnar
et al., 2009). Moreover, protracted brain maturation in adolescence is associated with increased cognitive control and
emotional maturity (e.g., Casey et al., 2008; Nelson et al.,
2005). For example, connections between the amygdala and
prefrontal as well as limbic brain structures that are important for the integration of emotion and cognition, are still
developing during adolescence (Cunningham et al., 2002).
Elevated cortisol levels in relation to an fMRI experiment
have indeed been shown in adolescents (Eatough et al., 2009).
However, the effect of this neuroendocrine stress response on
brain activation has not yet been studied in adolescents. The
first aim of the present study is to examine both biological
and subjective measures of experienced stress during an fMRI
experiment in young versus old male adolescents. We
selected homogenous groups of only male participants in
order to study the association between age, cortisol and brain
activation without the confounding influence of sex (and sexrelated hormones). Participants performed a motivated and
challenging gambling task in the scanner, a situation which
has been shown to activate the HPA-axis (Dickerson and
Kemeny, 2004; Kukolja et al., 2008). We predicted elevated
cortisol levels in response to the fMRI experiment, which
might be higher in old than young adolescents. The second
aim was to investigate the association of individual cortisol
levels with behavior and functional brain activation during
the gambling task. Based on previous findings, we expected a
negative association between cortisol output during scanning
and brain activation, especially in medial prefrontal areas.
2.
Results
2.1.
Salivary cortisol and subjective experienced anxiety
changes throughout an fMRI experiment
With regard to salivary cortisol, the within subjects 2-factor
repeated measures ANOVA (test day age group) revealed a
trend for a significant main effect of age group (F(1, 20) ¼ 4.31;
p¼0.051), with young adolescents showing lower cortisol levels
on all days than old adolescents. The main effect of test day
(F(1, 20) ¼ 2.46; p¼ 0.113) and the interaction effect of test day age
group (F(1, 20) ¼ 0.76; p¼0.482) were not significant, suggesting that
salivary cortisol levels were comparable between the practice
brain research 1598 (2015) 1–11
3
Fig. 1 – Cortisol changes during the fMRI experiment for each age group. Error bars indicate the SEM.
session day, the beginning of the fMRI test day and a random
control day at home (Fig. 1). A second repeated measures ANOVA
(measurement point on fMRI day age group) showed that
salivary cortisol levels significantly decreased during the course
of the fMRI experiment (F(1, 20) ¼6.42; p¼ 0.007; Fig. 1). Post hoc
comparisons showed that cortisol levels were lower after (measure 5) compared with during (measure 4; p¼0.020) and before
(measure 3; p¼0.005) the scan session. The decreasing cortisol
levels over time occurred irrespective of age group (interaction
effect F(1, 20) ¼0.11; p¼ 0.893). Additionally, a trend for a significant
main effect of age group (F(1, 20) ¼4.08; p¼0.057) was found, with
higher cortisol values in old compared with young adolescents.
To measure the total cortisol output during the fMRI experiment,
we calculated the Area Under the Curve (AUC) with respect to
ground (Pruessner et al., 2003). A trend for a significant main
effect of age group in the AUC value (T(21) ¼ 1.90; p¼ 0.07) was
found, with higher AUC values in old than in young male
adolescents. The general trend for higher cortisol values in old
compared with young adolescents might reflect a small difference in time of the scan session, which was scheduled somewhat earlier in the afternoon for old adolescents. Therefore, we
repeated all analyses with time of measure 3 as covariate, which
weakened the trend for age group differences in overall cortisol
values on separate days (p¼ 0.051), within the fMRI session
(p¼0.117) and for the AUC value (p¼0.160).
In the case of subjective experienced stress, 4 participants
(2 young and 2 old adolescents) had missing values for the
STAI state at the home measurement (measure 2). Regarding
the subjective experienced anxiety, a within subjects 2-factor
repeated measures ANOVA (test day age group) revealed a
trend for an interaction effect of test day and age group (F(1,
16) ¼ 3.57; p ¼0.054) and a significant main effect of test day
(F(1, 16) ¼3.68; p¼ 0.050; Fig. 2). Post-hoc analyses revealed that
anxiety scores differed between the training session, fMRI
test session and a random day at home for young adolescents
(F(1, 8) ¼6.16; p ¼0.029), but not for old adolescents (F(1, 8) ¼1.02;
p¼ 0.409). Young adolescents reported less anxiety at home
than during both the training day (p ¼0.017) and fMRI test
day (p¼ 0.047). Main effect of age group was not significant
(F(1, 20) ¼0.84; p¼ 0.373). A second repeated measures ANOVA
(measurement point on fMRI day age group) showed a trend
for a significant main effect of measurement point (F(1, 20) ¼
3.12; p ¼ 0.093), with lower reported anxiety after compared
with before the scan session (Fig. 2). Main effect of age group
(F(1, 20) ¼ 0.69; p ¼0.417) and interaction effect (F(1, 20) ¼0.65;
p¼ 0.431) were not significant. To measure the change in
reported anxiety with regard to the fMRI session we calculated the difference in STAI state score at measurement point
3 minus the mean STAI state scores at measurement points
1 and 2 (Muehlhan et al., 2011). To determine whether this
difference score was significant, an ANOVA repeated measures (measurement point age group) was performed. This
analysis revealed a significant main effect of measurement
point (F(1, 20) ¼ 4.43; p ¼0.048), with higher experienced anxiety
at the start of the fMRI test day compared with the control
days. The main effect of age group (F(1, 20) ¼ 0.49; p¼ 0.493) and
interaction effect of measurement point age group (F(1, 20) ¼
0.40; p ¼0.537) were not significant.
Overall, correlations between salivary cortisol and subjective
experienced anxiety determined at each separate measurement
point were non-significant. We did find a trend for a significant
correlation between the total amount of cortisol during the fMRI
session (AUC value) and the change in self-reported anxiety
between the control and the fMRI test day (r¼0.369; p¼0.091).
However, this association between cortisol AUC and difference
in STAI state score was stronger for old adolescents (r¼0.575;
p¼0.064) than for young adolescents (r¼0.354; p¼0.285).
2.2.
Behavioral performance on the gambling task and its
relation with cortisol
The behavioral results showed no age group differences in
mean reaction time (T(20) ¼ 0.48; p ¼ 0.634), reaction time variability (T(20) ¼ 0.17; p¼ 0.870) and total number of points
(T(20) ¼ 1.64; p ¼0.118). Age groups did differ on a so-called
contrast value indicating the distribution of gambling versus
4
brain research 1598 (2015) 1–11
passing responses (T(20) ¼ 2.17; p¼ p¼ 0.042). Younger adolescents gambled more in trials with low gains in which older
adolescents inhibited and gambled less in trials with high
gains in which older adolescents chose to gamble.
We used a multiple regression analysis to examine the
relation between individual cortisol output during the fMRI
experiment (AUC), age group and performance on the gambling task. None of the multiple regression models for the
behavioral measures on the gambling task, i.e., mean reaction time, reaction time variability, total number of points
and the contrast value, yielded any significant results.
2.3.
Relation between cortisol output, age and brain
activation
A multiple regression analysis was carried out to identify
specific brain regions in which activation amplitude during a
global gambling task contrast was modulated by individual
cortisol levels in young adolescents and in old adolescents
separately. Table 1 lists the clusters of voxels in which task
activation amplitude correlated significantly with cortisol output during the fMRI experiment as measured by the AUC value.
In young adolescents we found three clusters that survived the whole brain FWE cluster corrected significance level
of 0.05. They were located at respectively the anterior cingulate cortex, the posterior cingulate cortex and the parietooccipital sulcus (Table 1; Fig. 3A). In all three areas a negative
correlation between brain activation magnitude and cortisol
output was found.
In old adults, there were two significant clusters at the
FWE cluster corrected level, one in the inferior parietal cortex
and one in the superior precentral gyrus (Table 1; Fig. 3B). The
activation in these two areas was negatively correlated with
cortisol output in old adolescents
The interaction term of cortisol output (AUC) and age
group revealed no significant clusters at the FWE corrected
level. To further explore whether interaction effects corresponded to the different brain areas that revealed associations between brain activation and cortisol in young versus
old adolescents, we evaluated the interaction term at the
uncorrected voxel level as well (po0.001 combined with
Fig. 2 – STAI state changes during the fMRI experiment for each age group. Error bars indicate the SEM.
Table 1 – Results of the multiple regression analysis on task-related brain activation.
Brain area
BA
Talairach coordinates
x
(A) Effect of cortisol output (AUC) in young adolescents
r Anterior cingulate
32
3
l Parieto-occipital sulcus
31
24
r Posterior cingulate
30
6
(B) Effect of cortisol output in old adolescents
r Inferior parietal lob
40
36
r Superior precentral gyrus
6
12
Size
T
pcluster_corrected
Slope
y
z
48
57
51
12
21
15
25
17
17
7.50
6.22
5.04
0.003
0.025
0.025
( )
( )
( )
51
24
42
66
30
17
7.25
5.64
0.001
0.025
( )
( )
Note: In the multiple regression analysis the global task-related contrast images were used as dependent variable and cortisol as measured by
AUC, age group, interaction cortisol age group and time of cortisol measurement point 3 were used as predictors. Results were reported at
the FWE cluster corrected level. Size in number of voxels. BA¼ Brodmann area; r ¼right; l ¼left; ( ) ¼ negative correlation.
brain research 1598 (2015) 1–11
5
Fig. 3 – Clusters in which brain activation amplitude during task performance is associated with cortisol output during the
fMRI experiment as revealed by the multiple regression analysis in respectively young (A) and old (B) adolescents. All results
are visualized at the uncorrected level p¼ 0.001. 1 ¼ Anterior cingulate cluster (3, 48, 12); 2 ¼posterior cingulate cluster (6, 51,
15); 3¼ parieto-occipital sulcus cluster ( 24, 57, 21); 4 ¼inferior parietal cluster (36, 51, 42); 5 ¼ superior precentral gyrus
cluster (12, 24, 66).
cluster threshold Z5 voxels). At the uncorrected level, the
interaction term of cortisol output (AUC) and age group
revealed only four small clusters (located in the amygdala
(18, 3, 9), anterior cingulate (3, 48, 12 and 9, 30, 6) and
superior precentral gyrus ( 27, 27, 54)). The anterior cingulate clusters corresponded to the previously described
association between brain activation and cortisol reactivity
in young adolescents.
The majority of above-mentioned clusters showing a
significant association between brain activation and cortisol
output are located at the border of or close to negative
responding areas during performance of the gambling task.
Across age groups, deactivation during task execution was
found in several areas including the posterior cingulate
cortex, superior temporal, superior and medial frontal gyri,
which have been described as part of a default mode network
(e.g., Dosenbach et al., 2007; Greicius et al., 2003). In contrast,
task-related activation was seen in the anterior cingulate
cortex, anterior insula, inferior frontal sulcus, subcortical and
occipital areas, which have been described as part of a large
task-related network (e.g., Dosenbach et al., 2007; Sridharan
et al., 2008). Both task positive and task negative responding
areas correspond with the task related brain activation
pattern described in a previous study using this gambling
task (Keulers et al., 2011). One cortisol related significant
cluster, i.e., in the inferior parietal lob, is located at the border
of a task positive area. For the clusters that showed a main
effect of cortisol in either young or old adolescents (Table 1;
Fig. 3), percent signal change ranged from slightly positive or
zero for participants with low cortisol response to slightly/
moderate negative for participants with high cortisol output.
For all clusters significant at the FWE corrected level we
also examined the correlation between the strength of the
task-induced BOLD response (percent signal change) and the
subjective experienced stress level. Subjective stress experienced at the start of the fMRI session was indicated by the
difference in STAI state score on the fMRI day versus the
control days. In young adolescents, a negative correlation
between PSC and subjective stress was found in two clusters
(left parieto-occipital sulcus: r¼ 0.682, p ¼0.021; right posterior cingulate cortex: r¼ 0.743; p ¼0.009). In old adolescents, a
negative correlation was shown in the right superior precentral gyrus cluster (r¼ 0.666; p¼ 0.025).
3.
Discussion
The present study examined stress-related cortisol output in
relation to brain activation during an fMRI experiment in two
male adolescent age groups. Pre-imaging salivary cortisol
levels were elevated compared with cortisol levels during
and post-imaging in both age groups. The same decreasing
trend during the fMRI session was seen in subjective experienced anxiety, suggesting anticipatory stress regarding fMRI
scanning that decreased during the course of the scan
session. This is in line with previous studies showing a
reduction in anticipatory anxiety in adults undergoing (f)
MRI (Dantendorfer et al., 1997; Muehlhan et al., 2011). In
6
brain research 1598 (2015) 1–11
contrast, increased cortisol levels in response to the scanner
have previously been found in at least subgroups of participants, the so-called responders (Muehlhan et al., 2011).
Differences between the present findings and previous findings that showed increased cortisol levels after compared
with before the fMRI session, might be ascribed to differences
in sampling points during the fMRI session, time of day and
sample characteristics (e.g., Eatough et al., 2009; Tessner
et al., 2006). Although the decrease in cortisol during the
scan session might partially reflect the diurnal rhythm of
cortisol, this effect is minimized by scanning in the late
afternoon when the diurnal fluctuations are relatively small.
Moreover, the relative decrease in cortisol during the scan
session (43%) was larger compared to the relative cortisol
decrease over the same time interval in healthy adults,
described elsewhere (23%; Peeters et al., 2004).
In general, the association between salivary cortisol and
subjective experienced anxiety was non-significant at the
different measurement points. This is line with the literature
about stress in general (e.g., Campbell and Ehlert, 2012) and
MRI related stress specifically (e.g., Muehlhan et al., 2011).
This dissociation probably reflects different response systems
and moderating factors (Campbell and Ehlert, 2012), as we did
show a tendency to a significant correlation when the cortisol
output during the entire fMRI session was related to the
reported change in anxiety at the start of the fMRI session
(Schlotz et al., 2008). Cortisol is thought to be a more reliable
index than self-reports to measure stress associated with the
laboratory setting of an fMRI experiment, especially in developmental samples (Campbell and Ehlert, 2012; Nelson et al.,
2005).
Overall cortisol levels tended to be higher in old compared
with young adolescents. Although the age effect could partly
be attributed to a small difference in time of the scan session,
this finding corresponds with an age-related increase in basal
cortisol levels during adolescence (Adam, 2006). Likewise, the
neuroimaging results suggest an age-dependent effect of
cortisol on the brain. A negative correlation between cortisol
output and brain activation amplitude was found in all
clusters, conform previous findings (e.g., Dedovic et al.,
2009; Kukolja et al., 2008; Pruessner et al., 2008). The clusters
were located close to mainly negative responding areas
during task execution. Higher cortisol output during the fMRI
experiment was associated with stronger deactivation of
these brain areas. The exact location of the association
between cortisol and activation in the brain was, however,
dependent on age group.
In young adolescents, activation in the ventral anterior
cingulate cortex (BA 32) was negatively associated with
cortisol. This finding is in line with previous neuroimaging
studies showing deactivation of the medial PFC cortex as a
neural substrate of stress (Dedovic et al., 2009; Kern et al.,
2008; Pruessner et al., 2008; Ossewaarde et al., 2011). In
addition, the medial PFC is found to be underactive in
subjects with posttraumatic stress disorder (PTSD) (Etkin
and Wager, 2007). Furthermore, the anterior cingulate cluster
can be located in the anterior rostral division of the medial
PFC involved in emotional processing and mentalizing (e.g.,
Amodio and Frith, 2006). This is in line with the claim that the
medial PFC is part of the default mode network, which is
activated during rest and self-referential mental activity, but
deactivated during focused task execution (e.g., Greicius
et al., 2003; Gusnard and Raichle, 2001). The significant cluster
in young adolescents located in the posterior cingulate cortex
can be associated with the default mode network as well (e.g.,
Dosenbach et al., 2007; Greicius et al., 2003). Functional
connectivity studies have shown enhanced DMN connectivity
in stress-related psychiatric disorders, probably reflecting
increased self-referential mentalizing (Greicius et al., 2007;
Lanius et al., 2010). A cortisol-related increased deactivation
of the above-mentioned default mode areas might suggest
that participants with higher cortisol output engage in more
self-referential (stress-related) thoughts and therefore need
more effort to inhibit these processes for efficient task
execution (Philip et al., 2014). Nevertheless, inhibiting the
self-referential thoughts during task execution seems to be
successful given that increased deactivation is associated
with less task-unrelated thoughts (McKiernan et al., 2006).
Old adolescents also showed negative associations
between cortisol and brain activation, albeit in different brain
areas. The cluster in the inferior parietal lob was located at
the border of a positive responding area during task execution (Keulers et al., 2011). The right inferior parietal cortex has
indeed been described as part of both the dorsal attention
network (Corbetta and Shulman, 2002) and the task positive
network (Dosenbach et al., 2007; Sridharan et al., 2008). Our
finding might be explained by a disruption of ongoing
cognitive activities by the stress related to fMRI scanning,
which redirects attentional resources from task execution
towards stress-related internal thoughts (Seibert and Ellis,
1991; Philip et al., 2014). In line with this interpretation,
subjects with PTSD display attention deficits and attenuated
BOLD responses in attention-related parietal regions during
task performance (Pannu Hayes et al., 2009). The other cluster
revealed in old adolescents was located in the right superior
precentral gyrus. This location has been associated with the
default mode network in some studies (Mazoyer et al., 2001;
McKiernan et al., 2003), although not in others (Dosenbach
et al., 2007; Greicius et al., 2003) and specifically not in
children younger than 13 years of age (Thomason et al.,
2008). This developmental change in the default mode network might explain the age-dependent association in this
area, with old adolescents showing a negative association
between activation and cortisol, while this association was
absent in young adolescents.
In some of the clusters we showed that the brain activation associated with cortisol was also negatively correlated
with subjective experienced stress. This indicates that the
significant effects of cortisol on brain activation in both
adolescent groups can be attributed to the experienced stress
and anxiety during the fMRI experiment.
Aspects of the design should be taken into consideration
in interpreting the present results. Given large individual
variability in both HPA axis activity and brain activation
patterns, the relatively small sample of 22 participants may
not have provided adequate statistical power to reveal all
aspects of cortisol-dependent modulation of brain activity.
Nevertheless, this study did reveal some significant associations between cortisol and brain activation. Future research
should not only include larger sample sizes but also an
brain research 1598 (2015) 1–11
additional child and adult group to examine developmental
effects in stress output to fMRI scanning over a broader age
range. Furthermore, our sample was restricted to male
participants. Given sex-related differences in brain development (e.g., Giedd et al., 2006) as well as cortisol reactivity (e.g.,
Kajantie and Phillips, 2006), female adolescents should also
be included in future studies to clarify the influence of sex on
both cortisol and associations between cortisol and brain
activation in adolescents.
In summary, both young and old adolescents undergoing
fMRI showed anticipatory stress responses, as indicated by
elevated cortisol levels and subjective experienced anxiety
that decreased during the scan session. Cortisol output
during the fMRI experiment was in general negatively associated with brain activation. The exact location of this
association was, however, dependent on age group. These
findings suggest that cortisol output caused by the fMRI
experiment itself may cloud developmental fMRI findings in
adolescence. The present findings might therefore imply that
individual cortisol output during fMRI should be considered
as a confounder and integrated as a nuisance variable in
analyzing developmental changes in brain activation during
adolescence.
4.
Experimental procedures
4.1.
Participants and procedure
A subsample of participants from a larger study examining
developmental changes in brain activation during a gambling
task (Keulers et al., 2011) provided salivary cortisol measures.
A total of 24 participants from the seventh (12/13 year-olds)
and eleventh (16/17 year-olds) grades of pre-university education were included. All participants had normal or
corrected-to-normal vision, were free from psychiatric or
neurological abnormalities, were screened for MRI contraindications, and did not use medication that could influence
cognitive functioning or salivary cortisol. Written informed
consent was obtained from all participants and their parents.
All participants attended both a training session and a fMRI
scan session, which were scheduled with 7 days in between.
The training session in a mock scanner served to familiarize
the participants with the scanning environment and procedure and to provide practice doing the experimental task.
Furthermore, a number of neuropsychological tests, questionnaires and salivary cortisol were administered in the
training session (see for details Keulers et al., 2011). In the
fMRI session, adolescents performed different cognitive tasks
in the scanner, an anatomical scan was made and salivary
cortisol as well as the STAI were administered before, during
and after the scanning. Participants received travel expenses
and 25 euros. The study was approved by the ethical committee of the Faculty Psychology and Neuroscience of Maastricht University. Two participants were excluded due to
excessive head motion (for details see Section 4.3.3.2). This
resulted in a final sample of 22 right-handed males, divided
into two age groups: young adolescents (N¼11, mean
age¼ 12.9, SD¼ 0.3, range 12.3–13.4) and old adolescents
(N¼ 11, mean age¼16.8, SD¼ 0.3, range 16.3–17.3). Pubertal
7
status was measured with the self-report version of the
Pubertal Developmental Scale (PDS; Carskadon and Acebo,
1993, the original English version was translated into Dutch).
Of the young adolescents, four participants were rated prepubertal, five early pubertal and two midpubertal; of the old
adolescents, four participants were rated midpubertal, six
late pubertal and one subject postpubertal. Age groups did
not differ on estimated verbal intelligence (M¼114.86,
SD¼9.68, T(20) ¼ 0.24; p¼ 0.82), as assessed with the Peabody
Picture Vocabulary Test-III-NL (Dunn and Dunn, 2005), or on
estimated nonverbal intelligence (M¼ 122.45, SD¼4.30,
T(20) ¼ 0.59; p ¼0.56), as assessed with the Raven Standard
Progressive Matrices (Raven et al., 1998). The Child Behavior
Checklist (CBCL) and the Youth Self Report (YSR) (Achenbach
and Rescorla, 2001) were administered to screen for behavioral problems. All participants had scores on the total
problem scale of these questionnaires within 1SD of the
mean of a normative standardized sample. Furthermore,
age groups did not differ from each other on either the CBCL
total problem scale (T(20) ¼ 0.89; p ¼0.39) or the YSR total
problem scale (T(20) ¼ 0.43; p¼ 0.67). The Dutch version of
the State Trait Anxiety Inventory (STAI; Van der Ploeg et al.,
1980) Trait version was administered to measure trait anxiety. Age groups did not differ on STAI Trait scores
(T(20) ¼ 0.136; p¼ 0.89). Level of parental education was
inquired with a commonly-used Dutch education scale
(Directoraat-Generaal voor de Arbeidsvoorziening, 1989), as
higher levels of parental education have been associated with
higher performance on executive functioning tasks (Ardila
et al., 2005). Three parent-pairs (13.6%) had a moderate level
of education (elementary school to general secondary education) and 19 (86.4%) parent-pairs had a high level of education
(high vocational education to doctoral degree). Level of
parental education did not differ between age groups
(χ2(1) ¼ 0.39; p¼ 0.534). Thus, the two age groups used in the
present study were comparable concerning several background variables such as educational level, intelligence,
behavioral functioning, trait anxiety and parental level of
education.
4.2.
Measurements
4.2.1.
Salivary cortisol
A total of five saliva samples were obtained. The first two
samples were obtained at approximately the same time as
the start of the scan session on two preceding days: during
the practice session, after 45 min of explanation and neuropsychological testing (measure 1; mean time 16:06 (SD¼118
min)) and at home (measure 2; mean time 16:00 (SD¼ 87
min)). The third sample was collected at the start of the scan
session, after 15 min of practicing the tasks and preparing for
the scan session (measure 3; mean time 15:56 (SD¼86 min)).
After cortisol measure 3 participants entered the scanner and
performed the first functional run of the gambling task. After
this first run a fourth saliva sample was obtained in the
scanner (measure 4). Then the participant performed consecutively the first run of a gonogo task (not described here),
the second run of the gambling task, an anatomical scan, the
second run of a gonogo task (not described here), the third
run of the gambling task and a cognitive localizer (not
8
brain research 1598 (2015) 1–11
described here). The final sample was collected immediately
after the participant had finished the last functional run
(measure 5). Measures 4 and 5 were collected 13.7 (SD¼ 2.2)
and 76.5 (SD¼ 4.2) minutes, respectively, after the participant
had entered the bore of the magnet. All saliva samples were
collected in the mid and late afternoon to control for diurnal
fluctuations in cortisol level. There was a small but significant
difference in the start time of the scan session between
young (mean time 16:37 (SD¼52 min)) and old (mean time
15:15 (SD¼ 95 min)) adolescents (T(21) ¼2.50; p¼ 0.02). To rule
out possible confounding by this time difference, analyses
were performed with and without the time of cortisol measure 3 as a covariate.
Salivary cortisol was collected with a cotton dental roll,
stored in a plastic tube (Salivette; Sarstedt, Etten-Leur, the
Netherlands). Subjects were asked not to eat or smoke in the
1.5 h prior to saliva sampling. They were told to chew lightly
on the swab and to keep it fully in their mouth for 2 min. For
the cortisol sampling at home, subjects were instructed to
write down the exact time of the measurement on a form
that also contained the specific instructions mentioned
above. Samples were stored in subjects’ home freezers until
transport to the lab, where uncentrifuged samples were
frozen at 20 1C until analysis.
Salivary cortisol levels were determined in duplicate by
direct radioimmunoassay using 125I-cortisol and antiserum
made against the 3-CMO-BSA conjugate by Dr. J. Sulon,
University of Liege, Belgium. The lower detection limit of
the assay was 0.3 nmol/l, with mean intra-assay and interassay coefficients of variation less than 5% and 10%, respectively. All samples from an individual subject were analyzed
in the same assay to reduce sources of variability.
4.2.2.
State trait anxiety inventory
The Dutch version of the state trait anxiety inventory (STAI;
Van der Ploeg et al., 1980) was administered. The STAI is a
self-report questionnaire about the severity of perceived
anxiety, measuring both the temporary condition of state
anxiety and the general and long-standing condition of trait
anxiety. The STAI consists of 20 items addressing somatic,
affective, and cognitive aspects of anxiety, which have to be
rated by participants from ‘not at all’ (score 1) to ‘very much’
(score 4) applicable to them. Sum scores can range from 20 to
80. During the practice session we administered both the trait
and state version of the STAI, after 45 min of explanation and
neuropsychological testing. Additionally, the STAI state was
filled in repeatedly by participants together with the salivary
cortisol measurements, namely at a random day at home,
before and after the scan session on the fMRI test day.
4.2.3.
Gambling task
Participants performed a gambling task in which they could
either gamble or pass in order to earn as many points as
possible (Keulers et al., 2011; Mennes et al., 2008). In each trial
a horizontal bar divided into two colored parts was presented
(range from 5–95% to 50–50%), indicating the probability of an
imaginary token being hidden underneath. Participants could
guess under which part a token was hidden by pressing the
corresponding left or right button. Depending on the correctness of their choice, points could be won or lost. The points
that could be won were indicated above the bar (range 10–
100). The points that could be lost were presented below the
bar (range 0–100), with the most ambiguous proportions (50–
50%) coupled with the highest losses (80–100 points). Participants could also choose to pass by withholding their
response, resulting in a small reward of 20 points. All
participants started with 100 points. Feedback about the trial
and an update of their total score was provided in only 67% of
the trials, in order to disentangle the effects related to feedback processing from those related to the decision making
process.
The behavioral results from a previous study with this
task (Mennes et al., 2008) were used to delineate four trial
types along two dimensions. Task difficulty (exogenous vs
endogenous decision making) and task response (gambling vs
passing) were manipulated, leading to easy, exogenous trials
in which participants could either gamble (exo GAMBLE) or
pass (exo PASS) and to difficult, endogenous trials in which
participants could either gamble (endo GAMBLE) or pass
(endo PASS). These trial conditions were not relevant for
the current analyses. The task comprised three runs of
7.3 min and 50 trials each. The bar was presented for 3.5 s
and was, in 2/3 of the trials, followed 0.2–2.0 s later by
feedback for 1.7 s. The interval between the end of the
feedback and the start of a new trial varied between 1.0 and
4.0 s.
4.3.
Statistical analyses
4.3.1.
Cortisol and subjective experienced anxiety data
The statistical package SPSS 16.0 was used for the analyses.
Of all 110 salivary cortisol measures, measure 3 was missing
for one young adolescent and measure 5 was missing for
another young adolescent. These two missing cortisol values
were replaced by the mean cortisol value of the particular
measure for that age group (Tabachnick and Fidell, 2001).
Natural logarithmic transformations were applied before
analyses to meet the criteria for normality. To determine
cortisol changes over time and/or changes due to age, two
analyses of variance (ANOVA) were conducted. First, an
ANOVA was performed with salivary cortisol measured at
approximately the same time on separate test days (3 levels)
as within-subject variable and age group (2 levels) as
between-subject variable to examine differences in cortisol
values at the training session day (measure 1), a random day
at home (measure 2) and at the beginning of the scan session
day (measure 3). Second, an ANOVA was performed with
salivary cortisol measured different times within the scan
session (3 levels) as within-subject variable and age group (2
levels) as between-subject variable to examine changes in
cortisol values during the fMRI experiment. Furthermore, to
examine effects of cortisol on behavioral measures as well as
on brain activation we determined the Area Under the Curve
(AUC) with respect to ground for the cortisol values before,
during and after fMRI scanning using the trapezoid formulas
as suggested by Pruessner et al. (2003). This measure indicates the total cortisol output during the fMRI experiment.
Natural logarithmic transformations were applied to the AUC
values to meet criteria for normality. The advantage of the
AUC value is that the repeated measures are aggregated into
brain research 1598 (2015) 1–11
a single value, taking into account different time intervals
between measures. An independent sample t-test with age
group (2 levels) as between-subject variable was conducted to
examine age differences in AUC values.
To determine changes over time and/or changes due to age
regarding the STAI state score, two analyses of variance (ANOVA)
were conducted to determine changes in self-reported anxiety
between different days and different times within the fMRI test
session, respectively (analogous to the ANOVA’s for cortisol
data). Additionally, to examine associations of STAI state scores
with cortisol as well as with brain activation we calculated the
subjective stress experienced at the start of the fMRI session as
the difference in STAI state score at measurement point 3 minus
the mean STAI State scores at measurement points 1 and 2
(Muehlhan et al., 2011). To determine the significance of this
difference in STAI scores, an ANOVA with measurement point (2
levels: mean of measurement points 1 and 2 versus measurement point 3) as within-subject factor and age group (2 levels) as
between-subject variable.
Correlation analyses were performed to determine the
association between salivary cortisol and subjective experienced anxiety measures.
For all cortisol and subjective anxiety analyses, the alphavalue was set at 0.05 and trends were defined as 0.05opo0.10.
4.3.2.
Behavioral data on the gambling task
Independent samples t-tests were performed to examine age
group differences in behavioral performance on the gambling
task. Dependent variables were the mean reaction time,
reaction time variability, total number of points and a contrast value indicating the distribution of gambling versus
passing responses. This distribution was quantified as (MS)/
(SþM) where S is the proportion of gambling in trials with a
gain equal or lower than 20 points and M the proportion of
gambling in all other trials. This contrast value ranges
between 1 (only gambles in the first kind of trials) and 1
(only gambles in the other trials) (Mennes et al., 2009). Before
statistical analyses a cubic transformation was conducted on
the contrast measure to meet the criteria for normality.
To examine the influence of cortisol on task performance,
several multiple regression analyses were performed for the
different behavioral measures on the gambling task, i.e.,
mean reaction time, reaction time variability, total number
of points and the contrast value. The multiple regression
analyses included cortisol response indexed by the AUC
(continuous), age group (coded as: 0 ¼young and 1¼ old
adolescents) and the interaction between cortisol response
and age group. Time of cortisol measure 3 was included as
nuisance variable. The beta value associated with cortisol
response provided an estimate for the association between
cortisol and behavioral performance for young adolescents,
since the variables age group and the interaction were equal
to zero. In order to calculate the association between cortisol
response and behavioral performance in old adolescents, we
inverted the coding of the variable age group (0¼ old and
1¼ young adolescents). The assumptions of regression analysis (linearity, independence, homoscedasticity and normality distribution of the errors) were fulfilled for each model. For
all behavioral analyses, the alpha-value was set at 0.05 and
trends were defined as 0.05opo0.10.
9
4.3.3. Imaging data
4.3.3.1. Acquisition. A Siemens MAGNETOM Allegra 3T MRI
head-only scanner was employed. Head motion was constrained by the use of foam padding. A total of 32 axial slices
covering the whole brain including the cerebellum were imaged
by using a T2n-weighted gradient echo planner pulse sequence
(TR¼ 2000 ms, TE¼ 30 ms, FA¼90, FOV¼224, slice thickness¼ 4
mm, matrix size¼ 64 64, flip angle¼901). Voxel size was
3.5 3.5 4 mm. A T1-weighted anatomical scan was acquired
to aid with spatial normalization (TR¼ 2250 ms, TE¼ 2.6 ms, flip
angle¼ 91, FOV¼ 256 mm, slice thickness¼ 1 mm, matrix
size¼ 256 256, number of slices¼ 192). Voxel size was 1 1 1
mm. Slice scanning order was ascending interleaved.
4.3.3.2. Preprocessing. Data were preprocessed using BrainVoyager QX, version 1.9 (Brain Innovation, Maastricht, The
Netherlands). Images were corrected for slice scan time
differences, followed by rigid body motion correction, high-pass
temporal filtering, and spatial data smoothing using a Gaussian
kernel with a 4 mm full width at half maximum. Functional data
were coregistered with the anatomical volume using rigid body
transformation and manual adjustments to obtain optimal fit by
visual inspection if necessary. Subsequently, both functional and
structural volumes were transformed into standard stereotaxic
space using Talairach normalization.
To control for the confounding effect of head motion we
applied two corrections to the data (Johnstone et al., 2006).
First, head motion within one scan may distort the measured
fMRI signal by spatial misallocation. Therefore, we identified
scans during which head motion exceeded a particular threshold; absolute motion difference between 2 successive scans in z
direction greater than 0.4 mm (1/10th of the voxel size) and
rotation around dimension greater than 0.261 (angle corresponding to 0.4 mm z-displacement of frontopolar voxels,
assuming the rotation point in middle of brain is 88 mm from
the anterior end of the brain frontal pole (Talairach and
Tournoux, 1988). Task trials (taking into account the hemodynamic response function delay) cause signal intensity changes
with peak BOLD values 5 to 6 s after their occurrence. These
signal changes may overlap in time with head motion a few
scans later. Consequently, these signal changes are contaminated with within-scan head motion. Therefore, trials in a
preceding time window of 1 to 8 s of the identified scan were
modeled as an additional event of no interest. Because this
procedure reduces the number of trials available in the events
of interest, we excluded one young adolescent and one old
adolescent with more than 25% of their trials discarded. After
applying this procedure, the age groups did not differ on the two
most affected motion parameters for the remaining scans
(Mayer et al., 2007), namely translation in the z direction
(Tz; F(1, 20) ¼0.003; p¼ 0.959) and rotation in the x direction (Rx;
F(1, 20) ¼ 2.73; p¼0.114). Second, we included the six motion
parameters as covariates of no interest in our general linear
model to model signal intensity changes due to head motion
(Johnstone et al., 2006).
4.3.3.3. Analyses. Statistical analyses were performed on
individual participants’ data by using the general linear
model in SPM5 (Welcome Department of Cognitive Neurology, London, UK). The fMRI time series were modeled as
10
brain research 1598 (2015) 1–11
series of events convolved with a canonical hemodynamic
response function. A design matrix was set up to model all
task conditions of interest and the six motion parameters.
Error trials and trials contaminated with head motion were
modeled as two separate predictors, which were ignored in
the further analyses.
The global contrast images (all 4 task trial types pooled4fixation) computed on a subject-by-subject basis were taken to the
between-subject (second) level to examine correlations between
this activation pattern and cortisol, using a multiple regression
approach analogous to the behavioral data analyses (Kukolja
et al., 2008). The global task contrast was used as dependent
variable and the multiple regression included 4 predictors:
cortisol output indexed by the AUC, age group (coded as:
0¼young and 1¼ old adolescents), interaction between cortisol
response and age group and time of cortisol measure 3 as
nuisance variable. Comparable with the behavioral data, the
third predictor revealed the interaction between cortisol reaction
and age group, while the first predictor revealed an association of
cortisol output and brain activation in young adolescents. To
examine the association of cortisol and brain activation in old
adolescents the predictor age group was reversely coded
(1¼ young and 0¼ old adolescents). Significant associations were
identified at the FWE (po0.05) cluster corrected level combined
with a cluster threshold Z5 voxels.
references
Achenbach, T.M., Rescorla, L.A., 2001. Manual for the ASEBA
School-age Forms & Profiles. University of Vermont, Research
Centre for Children, Youth and Families, Burlington, VT.
Adam, E.K., 2006. Transactions among adolescent trait and state
emotion and diurnal andmomentary cortisol activity in
naturalistic settings. Psychoneuroendocrinology 31, 664–679.
Amodio, D.M., Frith, C.D., 2006. Meeting of minds: the medial
frontal cortex and social cognition. Nature 7, 268–277.
Ardila, A., Rosselli, M., Matute, E., Guajardo, S., 2005. The
influence of the parents’ educational level on the
development of executive functions. Dev. Neuropsychol. 28
(1), 539–560.
Buckert, M., Schwieren, C., Kudielka, B.M., Fiebach, C.J., 2014.
Acute stress affects risk taking but not ambiguity aversion.
Front. Neurosci. 8 (82), 1–11.
Campbell, J., Ehlert, U., 2012. Acute psychosocial stress: does the
emotional stress response correspond with physiological
responses? Psychoneuroendocrinology 37 (8), 1111–1134.
Carskadon, M.A., Acebo, C., 1993. A self-administered rating scale
for pubertal development. J. Adolesc. Health 14, 190–195.
Casey, B.J., Getz, S., Galvan, A., 2008. The adolescent brain. Dev.
Rev. 28, 62–77.
Corbetta, M., Shulman, G.L., 2002. Control of goal-directed and
stimulus-driven attention in the brain. Nat. Rev. Neurosci. 3,
201–215.
Cunningham, M.G., Bhattacharyya, S., Benes, F.M., 2002.
Amygdalo-cortical sprouting continues into early adulthood:
implications for the development of normal and abnormal
function during adolescence. J. Comp. Neurol 453, 116–130.
Dantendorfer, K., Amering, M., Bankier, A., Helbich, T., Prayer, D.,
Youssefzadeh, S., Alexandrowicz, R., Imhof, H., Katschnig, H.,
1997. A study of the effects of patient anxiety, perceptions,
and equipment on motion artifacts in magnetic resonance
imaging. Magn. Reson. Imaging 15, 301–306.
Dedovic, K., Duchesne, A., Andrews, J., Engert, V., Pruessner, J.C.,
2009. The brain and the stress axis: the neural correlates of
cortisol regulation in response to stress. Neuroimage 47,
864–871.
De Quervain, D.J., Roozendaal, B., Nitsch, R.M., McGaugh, J.L.,
Hock, C., 2000. Acute cortisone administration impairs
retrieval of long-term declarative memory in humans. Nat.
Neurosci. 3 (4), 313–314.
De Kloet, E.R., Oitzl, M.S., Joels, M., 1999. Stress and cognition: are
corticosteroids good or bad guys? Trends Neurosci 22 (10),
422–426.
Dickerson, S., Kemeny, M., 2004. Acute stressors and cortisol
responses: a theoretical integration and synthesis of
laboratory research. Psychol. Bull. 130 (3), 355–391.
Directoraat-Generaal voor de Arbeidsvoorziening. (1989).
Handleiding voor de functieanalyse [Function analyses
manual]. Den Haag: SDU uitgeverij.
Dosenbach, N.U.F., Fair, D.A., Miezin, F.M., Cohen, A.L., Wenger, K.
K., Dosenbach, R.A.T., Fox, M.D., Snyder, A.Z., Vincent, J.L.,
Raichle, M.E., Schlaggar, B.L., Petersen, S.E., 2007. Distinct
brain networks for adaptive and stable task control in
humans. Proc. Natl. Acad. Sci. U.S.A. 104 (26), 11073–11078.
Dunn, L.M., Dunn, L.M., 2005. Peabody Picture Vocabulary Test-IIINL. Handleiding [Manual]. Harcourt Assessment B.V.,
Amsterdam.
Eatough, E.M., Shirtcliff, E.A., Hanson, J.L., Pollak, E.A., 2009.
Hormonal reactivity to MRI scanning in adolescents.
Psychoneuroendocrinology 34, 1242–1246.
Etkin, A., Wager, T.D., 2007. Functional neuroimaging of anxiety: a
meta-analysis of emotional processing in PTSD, social anxiety
disorder, and specific phobia. Am. J. Psychiatry 164, 147–1688.
Giedd, J.N., Clasen, L.S., Lenroot, R., Greenstein, D., Wallace, G.L.,
Ordaz, S., et al., 2006. Puberty-related influences on brain
development. Mol. Cell. Endocrinol. 254–255, 154–162.
Greicius, M.D., Flores, B.H., Menon, V., Glover, G.H., Solvason, H.B.,
Kenna, H., Reiss, A.L., Schatzberg, A.F., 2007. Resting-state
functional connectivity in major depression: abnormally
increased contributions from subgenual cingulate cortex and
thalamus. Biol. Psychiatry 62, 429–437.
Greicius, M.D., Krasnow, B., Reiss, A.L., Menon, V., 2003.
Functional connectivity in the resting brain: a network
analysis of the default mode hypothesis. Proc. Natl. Acad. Sci.
U.S.A. 100 (1), 253–258.
Gusnard, D.A., Raichle, M.E., 2001. Searching for a baseline:
functional imaging and the resting human brain. Nat. Rev.
Neurosci. 2, 685–694.
Gunnar, M.R., Talge, N.M., Herrera, A., 2009. Stressor paradigms in
developmental studies: what does and does not work to
produce mean increases in salivary cortisol.
Psychoneuroendocrinology 34, 953–967.
Johnstone, T., Ores Walsh, K.S., Greischar, L.L., Alexander, A.L.,
Fox, A.S., Davidson, R., Oakes, T.R., 2006. Motion correction
and the use of motion covariates in multiple-subject fMRI
analysis. Hum. Brain Mapp 27, 779–788.
Kajantie, E., Phillips, D.I.W., 2006. The effects of sex and hormonal
status on the physiological response to acute psychosocial
stress. Psychoneuroendocrinology 31, 151–178.
Kern, S., Oakes, T.R., Stone, C.K., McAuliff, E.M., Kirschbaum, C.,
Davidson, R.J., 2008. Glucose metabolic changes in the
prefrontal cortex are associated with HPA axis response to a
psychosocial stressor. Psychoneuroendocrinology 33, 517–529.
Keulers, E.H.H., Stiers, P., Jolles, J., 2011. Developmental changes
between ages 13 and 21 years in the extent and magnitude of
the BOLD response during decision making. Neuroimage 54,
1442–1454.
Kirschbaum, C., Wolf, O.T., May, M., Wippich, W., Hellhammer, D.
H., 1996. Stress-and treatment-induced elevations of cortisol
brain research 1598 (2015) 1–11
levels associated with impaired declarative memory in
healthy adults. Life Sci. 58 (17), 1475–1483.
Kukolja, J., Thiel, C.M., Wolf, O.T., Fink, G.R., 2008. Increased
cortisol levels in cognitively challenged situations are
beneficial in young but not older subjects.
Psyhopharmacology 201, 293–304.
Lanius, R.A., Bluhm, R.L., Coupland, N.J., Hegadoren, K.M., Rowe,
B., Theberge, J., Neufeld, R.W., Williamson, P.C., Brimson, M.,
2010. Default mode network connectivity as a predictor of
post-traumatic stress disorder symptom severity in acutely
traumatized subjects. Acta Psychiatr. Scand. 121, 33–40.
Larson, R., Asmussen, L., 1991. Anger, worry, and hurt in early
adolescence: an enlarging world of negative emotions. In:
Colten, M.E., Gore, S. (Eds.), Adolescent Stress: Causes and
Consequences. Aldine de Gruyter, New York, NY, pp. 21–41.
Lupien, S.J., Maheu, F., Tu, M., Fiocco, A., Schramek, T.E., 2007. The
effects of stress and stress hormones on human cognition:
implications for the field of brain and cognition. Brain Cogn.
65, 209–237.
Lupien, S.J., Wilkinson, C.W., Briere, S., Menard, C., Kin, Ng Ying,
Nair, N.P., N.M., 2002. The modulatory effects of
corticosteroids on cognition: studies in young human
populations. Psychoneuroendocrinology 27 (3), 401–416.
Maheu, F.S., Joober, R., Lupien, S.J., 2005. Declarative memory
after stress in humans: differential involvement of the
β-adrenergic and corticosteroid systems. J. Clin. Endocrinol.
Metab. 90 (3), 1697–1704.
Mayer, A.R., Franco, A.R., Ling, J., Canive, J.M., 2007. Assessment
and quantification of head motion in neuropsychiatric
functional imaging research as applied to schizophrenia. J. Int.
Neuropsychol. Soc 13, 839–845.
Mazoyer, B., Zago, L., Mellet, E., Bricogne, S., Etard, O., Houde, O.,
Crivello, F., Joliot, M., Petit, L., Tzourio-Mazoyer, N., 2001.
Cortical networks for working memory and executive
functions sustain the conscious resting state in man. Brain
Res. Bull. 54, 287–298.
McKiernan, K.A., D’Angelo, B.R., Kaufman, J.N., Binder, J.R., 2006.
Interrupting the “stream of consciousness”: an fMRI
investigation. Neuroimage 29 (4), 1185–1191.
McKiernan, K.A., Kaufman, J.N., Kucera-Thompson, J., Binder, J.R.,
2003. A parametric manipulation of factors affecting taskinduced deactivation in functional neuroimaging. J. Cogn.
Neurosci 15, 394–408.
Melendez, J., McCrank, E., 1993. Anxiety-related reactions
associated with magnetic resonance imaging examinations.
JAMA 270, 745–747.
Mennes, M., Van den Bergh, B., Lagae, L., Stiers, P., 2009.
Developmental brain alterations in 17 year old boys are
related to antenatal maternal anxiety. Clin. Neurophysiol. 120,
1116–1122.
Mennes, M., Wouters, H., van den Bergh, B., Lagae, L., Stiers, P.,
2008. ERP correlates of complex human decision making in a
gambling paradigm: detection and resolution of conflict.
Psychophysiology 45, 714–720.
Muehlhan, M., Lueken, U., Wittchen, H., Kirschbaum, C., 2011.
The scanner as a stressor: evidence from subjective and
neuroendocrine stress parameters in the time course of a
functional magnetic resonance imaging session. Int. J.
Psychophysiol. 79, 118–126.
Nelson, E.E., Leibenluft, E., McClure, E.B., Pine, D.S., 2005. The
social re-orientation of adolescence: a neuroscience
11
perspective on the process and relation to psychopathology.
Psychol. Med. 35, 163–174.
Ossewaarde, L., Qin, S., Van Marle, H.J.F., van Wingen, G.A.,
Fernández, G., Hermans, E.J., 2011. Stress-induced reduction
in reward-related prefrontal cortex function. Neuroimage 55,
345–352.
Pannu Hayes, J., LaBar, K.S., Petty, C.M., McCarthy, G., Morey, R.A.,
2009. Alterations in the neural circuitry for emotion and
attention associated with posttraumatic stress
symptomatology. Psychiatry Res. Neuroimaging 172, 7–15.
Philip, N.S., Carpenter, S.L., Sweet, L.H., 2014. Developing
neuroimaging phenotypes of the default mode network in
PTSD: integrating the resting state, working memory, and
structural connectivity. J. Vis. Exp. 89, e51651.
Peeters, F., Nicolson, N.A., Berkhof, J., 2004. Levels and variability
of daily life cortisol secretion in major depression. Psychiatry
Res. 126, 1–13.
Pruessner, J.C., Dedovic, K., Khalili-Mahani, N., Engert, V.,
Pruessner, M., Buss, C., Renwick, R., Dagher, A., Meaney, M.J.,
Lupien, S., 2008. Deactivation of the limbic system during
acute psychosocial stress: evidence from positron emission
tomography and functional magnetic resonance imaging
studies. Biol. Psychiatry 63 (2), 234–240.
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.
Raven, J., Raven, J.C., Court, J.H., 1998. Manual for Raven’s
Progressive Matrices and Vocabulary Scales. Section 1, General
Overview. Harcourt Assessment, San Antonio, TX.
Schlotz, W., Kumsta, R., Layes, I., Entringer, S., Jones, A., Wüst, S.,
2008. Covariance between psychological and endocrine
responses to pharmacological challenge and psychosocial
stress: a question of timing. Psychosom. Med. 70 (7), 787–796.
Seibert, P.S., Ellis, H.C., 1991. Irrelevant thoughts, emotional mood
states, and cognitive task performance. Mem. Cogn. 19,
507–513.
Sridharan, D., Levitin, D.J., Menon, V., 2008. A critical role for the
right fronto-insular cortex in switching between centralexecutive and default-mode networks. Proc. Natl. Acad. Sci. U.
S.A. 105 (34), 12569–12574.
Tabachnick, B.G., Fidell, L.S., 2001. Using Multivariate Statistics.
Allyn and Bacon, Boston, London.
Talairach, J., Tournoux, P., 1988. Co-planar Stereotaxic Atlas of The
Human Brain. Thieme Medical Publishers Inc., New York, NY.
Tessner, K.D., Walker, E.F., Hochman, K., Hamann, S., 2006.
Cortisol responses of healthy volunteers undergoing magnetic
resonance imaging. Hum. Brain Mapp 27, 889–896.
Thomason, M.E., Chang, C.E., Glover, G.H., Gabrieli, J.D.E.,
Greicius, M.D., Gotlib, I.H., 2008. Default-mode function and
task-induced deactivation have overlapping brain substrates
in children. Neuroimage 41, 1493–1503.
Van der Ploeg, H.M., Defares, P.B., Spielberger, C.D., 1980.
Handleiding bij de Zelf Beoordelings Vragenlijst ZBV: Een
nederlandstalige bewerking van de Spielberger State-Trait
Anxiety Inventory, STAI-DY [Manual of the Self-Evaluation
Questionnaire: A Dutch Version of the State-Trait Anxiety
Inventory]. Swets & Zeitlinger B.V., Lisse, The Netherlands.
Weerda, R., Muehlhan, M., Wolf, O.T., Thiel, C.M., 2010. Effects of
acute psychosocial stress on working memory related brain
activity in men. Hum. Brain Mapp 31 (9), 1418–1429.