Psychoneuroendocrinology (2006) 31, 1087–1097
Serotonin regulation of the human stress response
Sean D. Hooda,Ã, Dana A. Hincea,b, Hayley Robinsona,Melita Cirilloa, David Christmasa, Joey M. Kayec
aSchool of Psychiatry and Clinical Neurosciences (M521), University of Western Australia,QEII Medical Centre, Perth, Nedlands, Western Australia 6009, AustraliabPsychopharmacology Unit, University of Bristol, Bristol, UKcDepartment of Endocrinology and Diabetes, Sir Charles Gairdner Hospital, Perth, Australia
Received 5 May 2006; received in revised form 6 July 2006; accepted 16 July 2006
Acute tryptophan depletion (ATD) is a technique that has been used to
evaluate the effects on humans of acutely reducing serotonin neurotransmission. We
have developed a model using a single breath of 35% CO
axis and produces autonomic and behavioural arousal, thus modelling a stress
response. This study combines ATD and single breath 35% CO2 inhalation to study
A randomised, double-blinded, placebo-controlled, cross-over trial involving 14
healthy adult volunteers aged between 18 and 65 years was undertaken. Subjectsunderwent double-blind tryptophan depletion over 2 days and were then crossedover 1 week later. During each study day, at the time of peak depletion, participantswere single blinded to receive a single breath of 35% CO2 or air. This was followed40 min later by the other gas. Psychological outcomes were assessed with theSpielberger State Anxiety Inventory (SSAI), Visual Analogue Scales (VAS), PanicInventory (PI), Panic and Agoraphobia Scale (PSI) and Beck Depression Inventory(BDI). Physiological outcome was measured by serial plasma cortisol, prolactin andtryptophan levels, pulse and blood pressure.
Tryptophan depletion did not exacerbate 35% CO2 inhalation effects on anxiety
symptoms. Single breath CO2 robustly increased plasma cortisol levels in comparisonto an air inhalation; this was less certain for prolactin levels. ATD influenced the HPAaxis (associated with higher cortisol levels), apparently independent of CO2 or airinhalation stressors. ATD and 35% CO2 inhalation both induced a pressor response andbradycardia in these normal volunteers.
Thirty-five percent CO2 inhalation and ATD independently activate the human
stress response, but do not appear to produce synergistic effects when combined, atleast for the conditions produced in this study. & 2006 Elsevier Ltd. All rights reserved.
ÃCorresponding author. Tel.: +61 8 9346 2393; fax: +61 8 9346 3828.
E-mail address: email@example.com (S.D. Hood).
0306-4530/$ - see front matter & 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.psyneuen.2006.07.001
dietary manipulation that causes a rapid reductionin plasma tryptophan levels (typically 70–80%
Stress leads to subjective anxiety, endocrine
activation (especially of the hypothalamo-pituitary
There have been several studies of the effect of
adrenal (HPA) axis) and cardiovascular changes
5-HT manipulation on CO2 challenge in both volun-
(Sinha et al., 1999). Existing tests of the human
teer and patient groups. One study (Miller et al.,
stress response are limited because they (a) are not
2000) that measured salivary cortisol reported no
reproducible in laboratory settings (b) are variable
changes in response to 5% CO2 inhalation after
in effect and involve painful stimuli or (c) are
tryptophan depletion, which our studies suggest is
confounded by other emotional effects. It is
too low a dose to increase cortisol (Kaye et al.,
therefore essential to develop a technique that
2004a). Two studies found that tryptophan depletion
may be employed to reliably and safely induce
resulted in more 35% CO2-induced increases in
stress responses in subjects in a controlled labora-
anxiety symptoms measured with the Panic Symp-
tom List Rating scale, in both volunteers (Klaassen et
The 35% carbon dioxide (CO2) single inhalation
al., 1998), and patients with panic disorder
paradigm fulfils these requirements (Kaye et al.,
(Schruers et al., 2000). Reducing brain 5-HT function
2004a). Firstly, it robustly activates the HPA axis
by another means, using the 5-HT antagonist
and produces a rapid and profound sympathetic
metergoline, also significantly enhanced the anxio-
activation as well as significantly increasing anxiety
genic effect of 35% CO2 (Ben Zion et al., 1999).
(Argyropoulos et al., 2002; van Duinen et al., 2005;
Taken together, the studies discussed above
Wetherell et al., 2006). For example, in the
suggest that agents that decrease serotonergic
Argyropoulos et al. study, 13 of 14 subjects showed
throughput may enhance CO2 stress responses in
an elevation of plasma cortisol in response to the
both patients and volunteers. However, the neu-
35% CO2 challenge, and salivary cortisol has
roendocrine and autonomic variables have not been
recently been shown to be elevated by 35% CO2
assessed under altered serotonergic conditions, and
inhalation (van Duinen et al., 2005). Also in the
an important question is whether manipulating
Argyropoulos et al. study, cardiovascular measure-
serotonergic function alters one or both of these
ments indicated autonomic arousal with immediate
parameters in this paradigm. The aim of the
activation of the sympathetic system demonstrated
present study, therefore, was to assess neuroendo-
by an immediate increase in blood pressure.
crine, autonomic and psychological responses to
Secondly, we have found that these effects show
single breath 35% CO2 inhalation in healthy volun-
good test–retest reliability over a period of several
teers after altering serotonergic function using the
weeks (Kaye et al., 2004a). And thirdly, the effects
tryptophan depletion technique. It was hypothe-
of CO2 on cortisol are dose related in that
sised that tryptophan depletion would increase the
inhalation of CO2 below a concentration of 25%
psychological (increase in subjective anxiety mea-
evokes little response, whereas at concentrations
sures), neuroendocrine (e.g. cortisol and prolactin
above 35%, glottal spasm is induced which in turn
increases) and autonomic (e.g. bradycardia and
impairs performance in the test. The 35% CO2
pressor response) responses to single breath 35%
challenge therefore appears to be the optimal
CO2 in normal volunteers. Prolactin was also
challenge in clinical research settings (Kaye et al.,
measured as it is known to be stress responsive
There are several lines of evidence suggesting a
role for the serotonin system in the stress response. Serotonergic neurones within the rostral medullamay have close associations with cerebral arteries
and function as primary chemosensory cells (Brad-ley et al., 2002). In addition, studies in animals
suggest that serotonergic projections to the peri-aqueductal grey region play an important role in
This was a randomized, double blind, placebo-
suppressing the behavioural and autonomic corre-
controlled, balanced cross-over study. Fourteen
lates of panic (Deakin and Graeff, 1991). Further,
healthy volunteers were tested on two similar
the hypothalamus and pituitary are densely inner-
days, 1 week apart. On the 1st day they were
vated by 5-HT neurons. It is possible to decrease
randomized in a double blind fashion into either
5-HT function by using the technique of tryptophan
tryptophan depletion (ATD) or non-tryptophan
depletion (acute tryptophan depletion (ATD)) (Bell
depletion (nTD) arms. At the time of peak deple-
et al., 2005; Hood et al., 2005). This involves a
tion subjects were challenged, in a single blind
fashion, with one forced breath of either 35% CO2,
diet the day before the test, and fasted from
or placebo (air). The other gas was administered in
midnight. On each test day the volunteers arrived
the same fashion 40 min later. Psychological out-
at the testing unit at 0900 h where they were rested
comes were measured at set times using standar-
in a semi-supine position on a bed. The testing room
was designed to be a non-stimulating environment
depression. Neuroendocrine outcomes were mea-
minimising excitement and stress. Subjects remained
sured by blood samples taken from an intravenous
in the test room for the duration of the experiment,
cannula that had been inserted at 0900 h to control
resting quietly, reading or performing similar activ-
for the effect of repeated venipuncture as a
ities, and were allowed to drink water for the
confounding stressor. Cardiovascular physiological
duration of the study. By 0930 h baseline ratings of
outcomes (blood pressure and heart rate) were
symptoms and cardiovascular measures were re-
measured using a non-invasive automated machine.
corded, and blood samples taken. Subjects then
The subjects were then crossed over into the other
consumed a tryptophan-free amino acid drink (the
arm of the study 1 week later for which their
ATD test) or a control drink (the nTD test) containing
tryptophan status remained double blinded and the
2.3 g of tryptophan. The other amino acids included
experiment continued to the same schedule. This
in both drinks (in the same proportion as human milk)
study was prospectively registered with the Inter-
are L-alanine, 5.5 g; L-arginine, 4.9 g; L-cysteine,
national Standard Randomised Controlled Trial
2.7 g; glycine, 3.2 g; L-histidine, 3.2 g; L-isoleucine,
Number Register (ISRCTN77876347). The protocol
8 g; L-leucine, 13.5 g; L-lysine, 11 g; L-methionine, 3 g;
was approved by the University of Western Aus-
tralia Human Research Ethics Committee.
6.9 g; L-threonine, 6.9 g; L-tyrosine, 6.9 g; and
L-valine 8.9 g. The order of the ATD and nTD testswas randomly allocated by unweighted blocking.
Women consumed only 80% of these drinks by virtueof their smaller body mass (Smith et al., 1997).
Volunteers were recruited via the UWA staff e-mail
Challenge testing was performed 5 h after consump-
database and by using personal contacts of the
tion of the TD or nTD drink, as this is the point of peak
research team. Informed consent was gained. Parti-
depletion (Hood et al., 2005). At the end of each test
cipants were screened by a medical practitioner for
day patients were given a meal and after having been
psychiatric and physical morbidity. This was achieved
assessed as recovered were allowed home. Subjects
using clinical interview, physical examination, a MINI
were given access to telephone contact over the
v5 semi-structured interview (Sheehan et al., 1998),
following 24 h and were discharged thereafter.
Panic and Agoraphobia Scale (PAS) (Bandelow, 1995),Panic Inventory (PI) (Coupland et al., 1997), BeckDepression Inventory (BDI) (Beck and Steer, 1987),
2.3.2. Thirty-five percent CO2 inhalation
Spielberger State (SSAI) and Spielberger Trait Anxiety
Inventories (STAI) (Spielberger et al., 1983). The
The CO2 provocation followed established proce-
Swedish Personality Questionnaire (SSP) (Gustavsson
dure (Argyropoulos et al., 2002). Gas delivery was
et al., 2000) and 24 h heart rate monitor (Stampfer,
single-blind, placebo controlled (air and 35% CO2).
1998) were additional screening measures. Exclusion
Order of delivery was randomised by unweighted
criteria were: age o18 or 465, a history of epilepsy,
blocking was the same for ATD and nTD occasions.
head injury, serious medical disorder, substance
The inhalation tests were performed at +5 and +5 h
abuse, current pregnancy or lactation, a psychiatric
40 min on both ATD and nTD days to minimise the
disorder within the last 6 months, arrhythmias,
effects of diurnal variability of neuroendocrine
tachycardia, hypertension and SSP results outside
measures and maximise the tryptophan depletion
one standard deviation from normal range. Preme-
effects. The gas mixtures were delivered via an oral
nopausal women were asked to present for testing in
gas regulator using the one vital capacity inhalation
the first 2 weeks of their menstrual cycles.
technique (van den Hout et al., 1987).
The tryptophan depletion procedure was performed
The three subjective ratings of anxiety were:
according to our standard protocol (Hood et al.,2005) based upon the established technique (Young
1. Visual Analogue rating Scales (VAS), measured on
et al., 1985). Subjects observed a low tryptophan
100 mm line, anchored from 0: ‘‘not at all’’ to
100: ‘‘the most y ever’’. Subjects were trained
Thayer, 1987), the Huyhn-Feldt epsilon (when
in the use of these scales (during screening and
Greenhouse-Geisser epsilon40.75) or the Green-
on the morning of each session) as integer
house-Geisser epsilon (otherwise) was used to
measures; these scales have been extensively
correct degrees of freedom when this assumption
used in similar settings (Bell et al., 2002;
was violated (Girden, 1992). A logarithmic trans-
formation was applied to cortisol and prolactin
2. The Panic Symptom Inventory (PSI): lists 34
data. The Bonferroni correction was applied to
symptoms related to a panic attack with the
post-hoc pair-wise tests. p-values less than 0.05
option of rating 0 ¼ not at all, 1 ¼ slight,
were considered statistically significant.
2 ¼ moderate, 3 ¼ severe, 4 ¼ very severe andhas been used in studies of anxiety provocation(e.g. Nutt et al., 1990).
3. The Spielberger State Anxiety Inventory (SSAI).
Each questionnaire was administered at baseline,
1.5 and 4 h post amino acid drink, 10 min prior to
The sample consisted of 14 subjects, of which nine
and 6 min post each gas inhalation, and at the end
were female. All the participants in this study were
of the study (30 min after eating a meal). For the
between the ages of 21 and 60 years with a mean
6 min post inhalation time point participants were
age of 34.5 years. The baseline variables shown in
instructed to complete the questionnaires accord-
Table 1 were the mean (SD) for the first time point
ing to how they felt at the peak effect of the
The SSP profile of this cohort showed all 13
parameters within 71 standard deviation of the
Serum prolactin was measured at baseline (0900 h),1400 h, 1 min before and 20 min after each gas
challenge. A further final measurement was takenat 1630 h, which was 30 min after eating a meal.
As shown in Fig. 2, the tryptophan depletion
Serum cortisol was measured 20 and 1 min prior
procedure reduced free tryptophan (fTRP) serum
as well as 20 min after each gas challenge. Base-
levels in the ATD condition, and increased fTRP
line, middle and endpoint values were also mea-
levels in the nTD condition. These observations
were supported by a significant depletion condition
Serum tryptophan was measured at 0900, 1510
by time interaction (F (2, 26) ¼ 79.39, po0.001).
There was no difference between the groups attime 0 (pre-drink) (F (1, 13) ¼ 0.11, p40.75).
At time 2 (5 h post amino acid drink and coinci-
Heart rate and blood pressure were measured at
À19, À5, À4, À3, À2, À1 before and +1, +2, +3,
had decreased by 68% from baseline under the
+4, +5, +19 min after each gas challenge. These
ATD condition, and had increased by 76% on
parameters were also measured at 0930, 1130,
the nTD occasion (F (1, 13) ¼ 192.2, po0.001). At
1402 and 1632 h to check variation throughout eachtest day.
Statistical analysis was conducted using SPSS v12.
Missing values in this dataset determined to be
missing-at-random were estimated using the multi-
ple imputation method using NORM software
(Schafer and Graham, 2002). Repeated measures
analysis of variance (ANOVA) was the primary
statistical technique used with time, depletion
status and gas inhalation type as within-subject
factors. As violation of the assumption of sphericityleads to increased type I error rates (Vasey and
Figure 1 Mean SSP personality profile for the sample (n ¼ 14). All are within 1 SD of the mean (À10 to +10). (mcg/mL)
Figure 2 Free tryptophan levels for acute tryptophan
depleted and non-depleted conditions at baseline (pre-
drink), 5 h post ingestion and after a meal. ***po0.001
nTD vs. ATD. Error bars represent the standard error of
PSI total score
the end of the study the fTRP levels between the
two conditions were still significantly different
(F (1,13) ¼ 187.7, po0.001). Similar results were
Figure 3 Total SSAI (A) and total PSI (B) scores pre and
found for total tryptophan levels (data not shown).
post gas inhalation under ATD and nTD conditions. ***po0.001 AIR vs. CO2, nTD condition. **po0.01, AIRvs. CO2, ATD condition. +++po0.001, AIR vs. CO2, ATDand nTD condition. Error bars represent the standard
To assess the impact of CO2 inhalation and
tryptophan depletion on psychological measures
Figure 3A displays the mean SSAI scores for the ATD
of anxiety, SSAI scores, PSI total scores, and VAS
and nTD days following AIR and CO2 inhalation. CO2
scores for the items ‘‘Anxiety’’, ‘‘Palpitations’’ and
inhalation significantly increased mean total SSAI
‘‘Something bad is just about to happen’’ were
scores compared with air inhalation (F (1, 13) ¼
analysed using time (pre- vs. post-inhalation),
25.12, po0.001) and there was a trend towards
depletion status (nTD vs. ATD) and gas (AIR vs.
higher SSAI scores in the ATD compared to the
CO2) as factors. Treatment orders and gender
nTD condition (F (1, 13) ¼ 3.89, p ¼ 0:07). How-
effects are only reported if statistically significant.
ever, this analysis failed to find evidence of an
interaction between gas inhalation and depletion
main effect of time (F (1, 13) ¼ 41.8, po0.001), of
status (F (1, 13) ¼ 0.30, p ¼ 0:595). That is, CO2
gas (F (1,13) ¼ 40.0, po0.001) and a significant gas
significantly increased mean SSAI scores to a similar
by time interaction (F (1,13) ¼ 39.0, po0.001). PSI
degree in both nTD (F (1, 13) ¼ 31.0, po0.001) and
scores increased post CO2 inhalation compared with
ATD (F (1, 13) ¼ 12.6, po0.01) conditions, com-
pre inhalation means in both the nTD (F (1, 13) ¼
A depletion order effect was found on total SSAI
po0.001) conditions but were not affected by AIR
scores, with participants depleted on the 1st day
inhalation in either of the depletion conditions. The
showing significantly lower SSAI scores compared to
main effect of depletion, and the interactions
those depleted on the 2nd day (ATD day 1:
involving depletion did not reach significance (all
M ¼ 28:7, 95% CI ¼ 25.3 to 32.1; ATD day 2:
p40.3). A similar pattern of results was seen in PSI
M ¼ 34:2, 95% CI ¼ 30.9 to 37.6; (F (1, 12) ¼ 6.4,
somatic and PSI psychological subscales (data not
po0.05). Inspection of the data found that this was
likely a consequence of four of the seven subjectsrandomised to the ATD condition on day 2 having
3.3.3. VAS-‘‘anxiety’’, ‘‘palpitations’’ and
STAI (trait) scores in the low clinical range (30 or
‘‘something bad is going to happen’’
above). To assess whether this failure of randomi-
Table 2 displays the mean VAS-anxiety scores pre-
sation had an impact on the pattern of effects seen
and post-inhalation for nTD and ATD conditions.
on the SSAI, the group was split according to normal
Analysis of the mean VAS scores revealed a
(20–30; n ¼ 9) or low clinical (31+; n ¼ 5) trait
significant gas by time interaction for VAS-anxiety
anxiety (STAI) scores and this factor was added
(F (1, 13) ¼ 25.6, po0.001), VAS-palpitations
to the ANOVA model. As expected, a main effect
(F (1, 13) ¼ 14.6, po 0.01) and VAS-‘‘something
of trait anxiety was seen on the SSAI means
bad is going to happen’’ (F (1, 13) ¼ 8.8, po0.05).
(F (1, 12) ¼ 5.1, po0.05), with people with high
All three measures increased significantly following
trait anxiety reporting more state anxiety. More
CO2 inhalation, but not after air inhalation (see
importantly, normal vs. low clinical state anxiety
Table 2). Tryptophan depletion had no effect on any
did not interact with any of the other factor (all
of these measures, as the main effect of and all
p’s40.1). Thus, it appears unlikely this order
interactions involving depletion status were not
effect can account for effects seen in the overall
3.3.2. PSIFigure 3B shows the mean PSI scores immediately
The mean cortisol levels, prolactin levels and
pre- and post-inhalation for both the ATD and nTD
cardiovascular measures across the test day are
conditions. Similar to that seen in the SSAI means,
displayed in Figs. 4 and 5. Baseline (t ¼ 0)
analysis of the mean total PSI scores measured
differences can be assumed not present unless
immediately pre and post gas inhalation revealed a
VAS mean (SD) for pre- and post-inhalation for each depletion status/inhalation condition. t rate (bpm) tisol (micr BASELINE PRL (micr t rate (1 min post-1 min pre) BASELINE
Figure 4 Cortisol (A) and prolactin (B) means as a
function of depletion status and inhalation challenge.
Figure 5 Top: Mean heart rate for ATD and nTD
*po0.05 pre 1 vs. post 20 in the CO2/ATD condition. Error
conditions for AIR and CO2 inhalation challenges. Bottom:
bars represent the standard error of the mean.
Distribution and mean (horizontal bar) of change in heartrate post inhalation challenge.
3.4.1. CortisolFigure 4A plays the mean cortisol levels for the nTDand ATD conditions at baseline, 20 and 1 min pre-,
po0.05). Post hoc ANOVA conducted at À20, À1
and 20 min post-, gas inhalation. Inspection of the
and +20 min separately found this interaction to be
graph demonstrates that cortisol decreased across
the consequence of a significant depletion status by
the day, in accordance with its usual circadian
gas interaction at post 20 min (F (1, 12) ¼ 9.5,
rhythm. The analysis of mean log cortisol levels 20
po0.01), but not at the other two time points.
and 1 min pre- and 20 min post-gas inhalation found
However, pair-wise comparisons between air and
cortisol levels to be significantly higher under the
CO2 conditions for nTD and ATD days were not
ATD condition compared with the nTD condition
(main effect of depletion status (F (1, 13) ¼ 4.7,po0.05)). The gas by time interaction (F (1.3,16.9) ¼ 5.5, po0.05) reflected the increase in
cortisol following CO2 inhalation, but not followingair inhalation, irrespective of depletion status. No
other main effects or interactions were significant.
Figure 5A displays the mean heart rate recordedbetween 19 min pre- to 19 min post-inhalation for
ATD and nTD conditions. Analysis of these means
Analysis of plasma prolactin levels pre (À20 and
found a near significant main effect of depletion
À1 min) and post (+20 min) gas inhalation for
status (F (1, 13) ¼ 4.4, po0.06), with higher heart
both depletion conditions. As anticipated, women
rates observed under the ATD condition. To assess
had higher levels of prolactin compared to men
the effect of the inhalation challenge on heart
(women: M ¼ 2:4, 95% CI ¼ 2.6–2.5; men: M ¼ 2:2,
rate, the difference between heart rate 1 min pre
95% CI ¼ 2.1–2.3; F (1, 12) ¼ 16.5, po0.01). Sex
and 1 min post breath was calculated. These
did not interact significantly with any other factor.
difference scores are seen in Fig. 5B. Analysis of
A trend towards a main effect of tryptophan
the difference scores revealed a trend towards a
main effect of gas (F (1, 13) ¼ 3.5, p ¼ 0:08), with
p ¼ 0:06), with prolactin levels tending to be higher
CO2 producing a greater reduction in heart rate
on the ATD day compared to the nTD day (see
compared to air inhalation. No other effects were
Fig. 4B). A significant depletion status by gas by
significant. One subject showed an increase in HR
time interaction was also seen (F (1.4, 16.3) ¼ 5.2,
by 14 beats/min in the nTD/CO2 condition. This was
(r ¼ 0:28, p40.3, n ¼ 14) or the nTD (r ¼ 0:12,
score ¼ 2.57). Removal of this participant from
the analysis rendered the main effect of challengesignificant (F (1, 12) ¼ 5.2, po0.05).
The aim of the present study was to assess
Figure 6A displays the mean systolic blood pressure
psychological, neuroendocrine and cardiovascular
recorded between 19 min pre- to 19 min post-
inhalation for ATD and nTD conditions. Analysis of
healthy volunteers after altering serotonergic
these means found no main effects or interactions
function. The tryptophan depletion procedure
to be significant. To assess the effect of the
produced a 68% reduction in free tryptophan levels.
inhalation challenge on systolic blood pressure,
This degree of reduction is congruent with reduc-
the difference between systolic blood pressure
tions reported in the literature that are associated
1 min pre- and 1 min post-breath was calculated
with increased anxiety in clinical populations (Bell
for each subjects under each condition. These
et al., 2005; Hood et al., 2005). Psychological,
difference scores are seen in Fig. 6B. Analysis of the
neuroendocrine and cardiovascular responses in
difference scores found no main effects or inter-
healthy volunteers were differentially affected by
actions to be significant. Furthermore, one partici-
ATD (no effect on self-reported anxiety in contrast
pant showed an extreme increase in SBP under the
with ATD-related increased in cortisol, prolactin
ATD/CO2 condition (standard score ¼ 2.75). Re-
and heart rate), but all measures were similarly
moval of this participant from the analysis did not
response was indicative of a stress response. Contrary to prediction, however, the present data
3.6. Cortisol reactivity and pressor response
provides very little evidence for a synergisticincrease in stress response in all three domains
To assess the relationship between changes in
cortisol and the pressor response post CO2, sepa-rate correlations were computed for these vari-
ables in the ATD and nTD condition. The change incortisol did not predict a significant amount of the
The 35% CO2 inhalation procedure (compared with
variance in the pressor response in either the ATD
air inhalation) was seen to be a reliable anxiogenictrigger as measured by SSAI, PSI and VAS scales, as
anticipated. Reduced serotonergic throughput via
tryptophan depletion (1) did not produce increasedanxiety ratings prior to CO2 challenge and (2) did
not augment self-rated anxiety produced by single
breath 35% CO2 inhalation in healthy volunteers.
The lack of synergistic effect between CO2 and
tryptophan depletion is unlikely to be a conse-
quence of a ceiling effect. The highest scores
recorded were at the most 2/3 of the maximum
score obtainable on all the psychological scales
These findings are consistent with previous
research. Firstly, studies that explicitly studied
the effect of ATD on anxiety in normal volunteers
report that ATD alone had little effect on anxiety
ratings (Goddard et al., 1995; Klaassen et al., 1998;
Miller et al., 2000), albeit small increases in
SBP (1 min post - 1 min pre)
‘‘nervousness’’ were reported on VAS (Goddard
et al., 1995) and STAI (Klaassen et al., 1998)measures. Two positive studies reported minor
Figure 6 Top: Mean systolic blood pressure for ATD andnTD conditions for AIR and CO
increases in anxiety ratings in male subjects who
Bottom: Distribution and mean (horizontal bar) of change
also reported an increase in depressive symptoma-
in systolic blood pressure post inhalation challenge.
tology (Smith et al., 1987; Benkelfat et al., 1994);
other studies that included anxiety ratings in
the activity of the HPA axis. Serotonin is known to
influence the HPA axis through direct actions at the
et al., 1985; Abbott et al., 1992; Park et al.,
hypothalamic, pituitary and adrenal levels (Chaoul-
1994; Cleare and Bond, 1995; Oldman et al., 1995;
off, 2000) with serotonergic systems having both
Ellenbogen et al., 1996; Smith et al., 1997; Barr et
facilitatory and inhibitory actions on the axis
al., 1997; Moore et al., 1998). Secondly, Klaassen
(Lowry et al., 2005). Further, these systems may
and colleagues found that volunteers exposed to
also act through the hippocampal negative feed-
single-blind 35% CO2 exhibited higher scores on the
back regulation of the HPA axis causing disinhibition
Panic Symptom Inventory on ATD vs. control
of the axis and a rise in cortisol levels (Lowry,
occasions; no ATD by 35% CO2 effect was seen on
2002). Despite the higher baseline cortisol levels on
a VAS-anxiety item (Klaassen et al., 1998). The
the ATD day, and the possibility of HPA axis
same group did, however, demonstrate a signifi-
disinhibition, the cortisol rise following CO2 was
cantly greater anxiety response to the CO2 chal-
not enhanced as might be anticipated under these
lenge in tryptophan depleted panic disorder
circumstances. This is also consistent with our
patients compared with patients who were trypto-
findings in volunteers with both physiological
phan replete (Schruers et al., 2000). Of interest,
(during lactation) and pharmacological (naltrexone
premedication of healthy volunteers with the
premedication) disinhibition of the HPA axis. In
known anxiogenic agents CCK-4 (Pols et al., 1999)
both circumstances, baseline cortisol levels were
or yohimbine (Pols et al., 1994) had no synergistic
significantly greater than control or placebo condi-
effect on anxiety response following the 35% CO2
tions, but the increase in cortisol following 35% CO2
challenge. No notable interaction between ATD and
was no greater (Kaye et al., 2004b; Kaye, 2005).
5% CO2 inhalation was seen in another group of
It is not clear why ATD produced a neuroendo-
volunteers (Miller et al., 2000), although this may
crine stress response in volunteers in the absence
reflect an inadequate CO2 concentration (Kaye
of a psychological stress response. This finding
demonstrates that, although often co-occurring,the different dimensions of the stress response donot necessarily need to appear together. This
highlights the importance of assessing the stressresponse in all 3 domains.
Single breath 35% CO2 robustly increased plasmacortisol levels in comparison with air inhalation. These results are congruent with previous findings
in healthy volunteers and indicates HPA axisactivation (Argyropoulos et al., 2002; van Duinen
The results presented here go someway to replicate
et al., 2005; Wetherell et al., 2006). Tryptophan
previous work concerning 35% CO2 effects upon
depletion did not facilitate this effect, but pro-
heart rate and blood pressure (Argyropoulos et al.,
duced an increase in cortisol levels compared to
2002; Kaye et al., 2004a, 2006; Wetherell et al.,
the nTD condition. Although not statistically sig-
2006). CO2 inhalation produced significant brady-
nificant pre inhalation, the consistently higher
cardia compared with air inhalation, consistent
cortisol levels across time in this group became
with previous findings. This was not influenced by
significant post inhalation. This is in keeping with
tryptophan depletion or related to plasma cortisol
the known mechanism of tryptophan depletion,
levels and is likely a parasympathetic response
that is, central 5-HT gradually decreases with time,
discrete from either sympathetic, HPA or arousal
and is maximally effective near the time of gas
centres (Kaye et al., 2006; Wetherell et al., 2006).
Although visual inspection of the data suggested
It is of some interest that ATD itself acted as a
that systolic blood pressure was increased by the
stressor, as measured by plasma cortisol (and to a
35% CO2 inhalation (and perhaps air inhalation
lesser extent prolactin) increases. It may be that
under ATD conditions: see Fig. 6), we failed to find
ATD produces an aversive state that is perceived as
these effects to be statistically significant. This
stressful and indirectly results in increased cortisol
may be a consequence of limited temporal grada-
levels. We found little evidence for this however, as
tion of the equipment used to measure blood
participants did not report the ATD day as being
pressure. The results of Kaye et al. (2006) suggest
significantly more stressful or aversive than the nTD
that the maximum pressor response is approxi-
condition on the basis of self-report measures.
mately 45 s post CO2 inhalation—we were only
Another explanation might be that reduced central
able to measure blood pressure every minute.
5-HT via tryptophan depletion may be modulating
Furthermore, Kaye et al. considered the maximum
change in blood pressure post breath, which takes
Argyropoulos, S.V., Bailey, J.E., Hood, S.D., Kendrick, A.H., Rich,
into account individual differences in this response,
A.S., Laszlo, G., Nash, J.R., Lightman, S.L., Nutt, D.J., 2002.
thereby increasing the power to detect an effect.
Inhalation of 35% CO2 results in activation of the HPA axis inhealthy volunteers. Psychoneuroendocrinology 27, 715–729.
This was only possible at 1 min intervals in the
Bandelow, B., 1995. Assessing the efficacy of treatments for
present study, and therefore it may still be the case
panic disorder and agoraphobia. II. The Panic and Agorapho-
that the maximum response for an individual was
bia Scale. Int. Clin. Psychopharmacol. 10, 73–81.
not detected. Therefore, future work should
Barr, L.C., Heninger, G.R., Goodman, W., Charney, D.S., Price,
ideally use a beat-to-beat measurement of blood
L.H., 1997. Effects of fluoxetine administration on mood
response to tryptophan depletion in healthy subjects. Biol. Psychiatry 41, 949–954.
The trend towards higher blood pressures seen on
Beck, A.T., Steer, R.A., 1987. Beck Depression Inventory Manual.
the ATD day suggests a modulatory role for
The Psychological Corporation, Harcourt Brace Jovanovic,
serotonin on the stress-induced pressor response.
There is some recent evidence that a reduction in
Bell, C., Forshall, S., Adrover, M., Nash, J., Hood, S.,
central serotonin by tryptophan depletion is asso-
Argyropoulos, S., Rich, A., Nutt, D., 2002. Does 5-HT restrainpanic? A tryptophan depletion study in panic disorder
ciated with significant excess pressor response in
patients recovered on paroxetine. J. Pharmacol. 16, 5–14.
anxious patients given a stress challenge (Davies
Bell, C., Hood, S.D., Nutt, D.J., 2005. Acute tryptophan
depletion. Part II: clinical effects and implications. Aust. N.
One limitation of the present study was that the
capacity of the inhalation breath was not mea-
Ben Zion, I.Z., Meiri, G., Greenberg, B.D., Murphy, D.L.,
sured. It is possible also that the reduced effect of
Benjamin, J., 1999. Enhancement of CO2-induced anxiety inhealthy volunteers with the serotonin antagonist metergo-
CO2 on blood pressure is a consequence of
line. Am. J. Psychiatry 156, 1635–1637.
incomplete inhalations, leading to a reduced CO2
Benkelfat, C., Ellenbogen, M.A., Dean, P., Palmour, R.M., Young,
‘‘dose’’. Although possible, this explanation is
S.N., 1994. Mood-lowering effect of tryptophan depletion.
unlikely as cortisol levels increased following CO
Enhanced susceptibility in young men at genetic risk for
major affective disorders. Arch. Gen. Psychiatry 51, 687–697.
Bradley, S.R., Pieribone, V.A., Wang, W., Severson, C.A., Jacobs,
An explanation of the differential effects of ATD on
R.A., Richerson, G.B., 2002. Chemosensitive serotonergic
patient groups but not on healthy volunteers might
neurons are closely associated with large medullary arteries.
be that patients have a vulnerability to serotonergic
manipulation that healthy populations do not. The
established finding of SSRI effectiveness for these
patient groups could therefore be viewed as a means
Cleare, A.J., Bond, A.J., 1995. The effect of tryptophan
depletion and enhancement on subjective and behavioural
of promoting serotonergic resiliency. Further delinea-
aggression in normal male subjects. Psychopharmacology
tion of the ways by which serotonergic systems may
modulate the human stress response is an active area
Coupland, N.J., Lillywhite, A., Bell, C.E., Potokar, J.P., Nutt,
D.J., 1997. A pilot controlled study of the effects offlumazenil in posttraumatic stress disorder. Biol. Psychiatry41, 988–990.
Davies, S.J., Hood, S.D., Argyropoulos, S.V., Morris, K., Bell, C.,
Witchel, H.J., Jackson, P.R., Nutt, D.J., Potokar, J., 2006. Depleting serotonin enhances both cardiovascular and psy-
Unlike studies of tryptophan depletion in patient
chological stress reactivity in recovered patients with anxietydisorders. J. Clin. Psychopharmacol. 26 (4), 414–418.
groups (e.g. panic disorder, depression, etc.), ATD
Deakin, J.F., Graeff, F.G., 1991. 5HT and mechanisms of defence.
did not augment the psychological response to a 35%
CO2 stressor in healthy controls. Furthermore, there
Ellenbogen, M.A., Young, S.N., Dean, P., Palmour, R.M.,
appears to be no additive effect of ATD on endocrine
Benkelfat, C., 1996. Mood response to acute tryptophan
or cardiovascular responses to this challenge. A novel
depletion in healthy volunteers: sex differences and tempor-al stability. Neuropsychopharmacology 15, 465–474.
finding from the present study was that ATD was
Freeman, M.E., Kanyicska, B., Lerant, A., Nagy, G., 2000.
itself a neuroendocrine stressor, the mechanism of
Prolactin: structure, function, and regulation of secretion.
which requires further investigation.
Girden, E.R., 1992. ANOVA: Repeated Measures. Sage, Newbury
Goddard, A.W., Charney, D.S., Germine, M., Woods, S.W.,
Heninger, G.R., Krystal, J.H., Goodman, W.K., Price, L.H.,1995. Effects of tryptophan depletion on responses to
Abbott, F.V., Etienne, P., Franklin, K.B., Morgan, M.J., Sewitch,
yohimbine in healthy human subjects. Biol. Psychiatry 38,
M.J., Young, S.N., 1992. Acute tryptophan depletion blocks
morphine analgesia in the cold-pressor test in humans.
Gustavsson, J., Bergman, H., Edman, G., Ekselius, L., von
Psychopharmacology (Berl) 108, 60–66.
Knorring, L., Linder, J., 2000. Swedish universities Scales of
Personality (SSP): construction, internal consistency and
Pols, H., Griez, E., Verburg, K., van der, W.D., 1994.
normative data. Acta Psychiatr. Scand. 102, 217–225.
Yohimbine premedication and 35% CO2 vulnerability in
Hood, S.D., Bell, C., Nutt, D.J., 2005. Acute tryptophan
healthy volunteers. Eur. Arch. Psychiatry Clin. Neurosci.
depletion. Part I: rationale and methodology. Aust. N. Z.
Schafer, J.L., Graham, J.W., 2002. Missing data: our view of the
Kaye, J., 2005. Mechanisms and clinical implications of the
state of the art. Psychol. Methods 7, 147–177.
neuroendocrine response to a novel carbon dioxide stressor in
Schruers, K., Klaassen, T., Pols, H., Overbeek, T., Deutz, N.E.,
man. Ph.D. Thesis, The University of Western Australia.
Griez, E., 2000. Effects of tryptophan depletion on carbon
Kaye, J., Buchanan, F., Kendrick, A., Johnson, P., Lowry, C.,
dioxide provoked panic in panic disorder patients. Psychiatry
Bailey, J., Nutt, D., Lightman, S., 2004a. Acute carbon
dioxide exposure in healthy adults: evaluation of a novel
Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs,
means of investigating the stress response. J. Neuroendocri-
J., Weiller, E., Hergueta, T., Baker, R., Dunbar, G.C., 1998.
The Mini-International Neuropsychiatric Interview (M.I.N.I.):
Kaye, J., Soothill, P., Hunt, M., Lightman, S., 2004b. Responses
the development and validation of a structured diagnostic
to the 35% CO challenge in postpartum women. Clin.
psychiatric interview for DSM-IV and ICD-10. J. Clin.
Kaye, J.M., Young, T.M., Mathias, C.J., Watson, L., Lightman,
Sinha, S.S., Coplan, J.D., Pine, D.S., Martinez, J.A., Klein, D.F.,
S.L., 2006. Neuroendocrine and behavioural responses to CO2
Gorman, J.M., 1999. Panic induced by carbon dioxide
inhalation in central versus peripheral autonomic failure.
inhalation and lack of hypothalamic–pituitary–adrenal axis
activation. Psychiatry Res. 86, 93–98.
Klaassen, T., Klumperbeek, J., Deutz, N.E., van Praag, H.M.,
Smith, K.A., Fairburn, C.G., Cowen, P.J., 1997. Relapse of
Griez, E., 1998. Effects of tryptophan depletion on anxiety
depression after rapid depletion of tryptophan. Lancet 349,
and on panic provoked by carbon dioxide challenge.
Smith, S.E., Pihl, R.O., Young, S.N., Ervin, F.R., 1987. A test of
Lowry, C.A., 2002. Functional subsets of serotonergic neurones:
possible cognitive and environmental influences on the mood
implications for control of the hypothalamic–pituitary–adre-
lowering effect of tryptophan depletion in normal males.
nal axis. J. Neuroendocrinol. 14, 911–923.
Psychopharmacology (Berl) 91, 451–457.
Lowry, C.A., Johnson, P.L., Hay-Schmidt, A., Mikkelsen, J.,
Spielberger, C.D., Gorsuch, R.L., Lushene, R., Vagg, P.R., Jacobs,
Shekhar, A., 2005. Modulation of anxiety circuits by seroto-
G.A., 1983. Manual for the State-Trait Anxiety Inventory.
nergic systems. Stress 8, 233–246.
Consulting Psychologists’ Press, Palo Alto, CA.
Miller, H.E., Deakin, J.F., Anderson, I.M., 2000. Effect of acute
Stampfer, H.G., 1998. The relationship between psychiatric
tryptophan depletion on CO2-induced anxiety in patients with
illness and the circadian pattern of heart rate. Aust. N. Z.
panic disorder and normal volunteers. Br. J. Psychiatry 176,
van den Hout, M.A., van der Molen, G.M., Griez, E., Lousberg,
Moore, P., Gillin, C., Bhatti, T., DeModena, A., Seifritz, E., Clark,
H., Nansen, A., 1987. Reduction of CO2-induced anxiety in
C., Stahl, S., Rapaport, M., Kelsoe, J., 1998. Rapid
patients with panic attacks after repeated CO2 exposure. Am.
tryptophan depletion, sleep electroencephalogram, and
mood in men with remitted depression on serotonin reuptake
van Duinen, M.A., Schruers, K.R., Maes, M., Griez, E.J., 2005.
inhibitors. Arch. Gen. Psychiatry 55, 534–539.
CO2 challenge results in hypothalamic–pituitary–adrenal
Nutt, D.J., Glue, P., Lawson, C., Wilson, S., 1990. Flumazenil
activation in healthy volunteers. J. Psychopharmacol. 19,
provocation of panic attacks. Evidence for altered benzodia-
zepine receptor sensitivity in panic disorder. Arch. Gen.
Vasey, M.W., Thayer, J.F., 1987. The continuing problem
of false positives in repeated measures ANOVA in psychophy-
Oldman, A., Walsh, A., Salkovskis, P., Fairburn, C.G., Cowen,
P.J., 1995. Biochemical and behavioural effects of acute
tryptophan depletion in abstinent bulimic subjects: a pilot
Wetherell, M.A., Crown, A.L., Lightman, S.L., Miles, J.N., Kaye,
study. Psychol. Med. 25, 995–1001.
J., Vedhara, K., 2006. The four-dimensional stress test:
Park, S.B., Coull, J.T., McShane, R.H., Young, A.H., Sahakian,
psychological, sympathetic–adrenal–medullary, parasympa-
B.J., Robbins, T.W., Cowen, P.J., 1994. Tryptophan depletion
thetic and hypothalamic-pituitary-adrenal responses follow-
in normal volunteers produces selective impairments in
ing inhalation of 35% CO2. Psychoneuroendocrinology 31,
learning and memory. Neuropharmacology 33, 575–588.
Pols, H., Griez, E., Bourin, M., Schruers, K., 1999. Effect of CCK-4
Young, S.N., Smith, S.E., Pihl, R.O., Ervin, F.R., 1985. Trypto-
on a 35% carbon dioxide challenge in healthy volunteers. Prog.
phan depletion causes a rapid lowering of mood in normal
Neuropsychopharmacol. Biol. Psychiatry 23, 1345–1350.
males. Psychopharmacology (Berlin) 87, 173–177.
ABRIL NEGADO E SEMPRE RENOVADO Comemora-se a data de 25 de Abril de 1974, negando Abril. Mesmo muitos dos que viveram e recordam o espírito do Abril popular fazem-no agora mais como romagem de saudade aos belos tempos do PREC, do que pela afirmação concreta e convicta, nos dias de hoje, da luta pelos sonhos de então. No discurso oficial repetem-se as bonitas e velhas palavras gastas para