This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Available online at www.sciencedirect.com
Biological Psychology 78 (2008) 87–92
www.elsevier.com/locate/biopsycho
To inhale or not to inhale: Conditioned avoidance in breathing
behavior in an odor—20% CO2 paradigm
Stien Fannes, Ilse Van Diest *, Ann Meulders, Steven De Peuter,
Debora Vansteenwegen, Omer Van den Bergh
Department of Psychology, University of Leuven, Tiensestraat 102, 3000 Leuven, Belgium
Received 22 June 2007; accepted 20 January 2008
Available online 31 January 2008
Abstract
This study investigated breathing behavior in an odor—CO2-inhalation fear conditioning paradigm. A differential conditioning paradigm was
applied in 55 participants. Both acquisition and extinction consisted of three CS+ and three CS trials. Diluted ammonia and butyric acid served as
conditional odor cues (CSs); inhalation of 20% CO2-enriched as US. The US was presented 10 s after CS+ onset and both stimuli co-terminated
30 s later. Subjective anxiety and US-expectancy were measured online upon presentation of the CSs. Respiratory behavior showed a biphasic
pattern during CS+ acquisition trials. Participants paradoxically lowered their ventilation first; an increased ventilation was observed only towards
the end of the trial. Extinction of this breathing inhibition was found. Participants avoiding the CO2 during acquisition did not show a reduction in
fear from acquisition to extinction, whereas Non-avoiders did. We conclude that paradoxical decreases in ventilation constitute a relevant
behavioral index of fear in CO2-inhalation paradigms.
# 2008 Elsevier B.V. All rights reserved.
Keywords: Respiration; Avoidance; Conditioning; CO2-inhalation; Odors; Fear
Inhalation of CO2-enriched air has been used repeatedly as
an unconditional stimulus (US) in human conditioning fear
paradigms (e.g., Acheson et al., 2007; Devriese et al., 2006;
Forsyth and Eifert, 1998; Forsyth et al., 1996). The rationale to
use CO2 rather than more traditional, exteroceptive USs, such
as electrocutaneous stimuli or white noise, is that CO2 is more
suitable as an analogue of the response patterns characteristic
for patients diagnosed with anxiety disorders, especially panic
(Forsyth et al., 1996; Griez et al., 2007). Depending on the
concentration and the duration, inhalation of CO2-enriched air
causes escalating symptoms of arousal, mimicking to a certain
extent the topography of human fear responses: increased
breathing, dizziness, a sense of dyspnea (breathlessness),
increased heart rate, reactive hyperemia, sweaty palms, feeling
of unreality, etc. As such, CO2-inhalation can be applied to
experimentally induce a ‘‘false alarm’’, i.c., an abrupt
autonomic activation in the absence of real threat or harm
(Barlow, 1988; Forsyth and Eifert, 1996). Pairing such a false
alarm (CO2-inhalation) with internal/external neutral cues
* Corresponding author. Tel.: +32 16 32 60 29; fax: +32 16 32 59 23.
E-mail address: [email protected] (I. Van Diest).
0301-0511/$ – see front matter # 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.biopsycho.2008.01.003
(conditional stimuli, CSs) in the laboratory may produce socalled ‘‘learned alarms’’, i.c., autonomic activation and
subjective fear in response to these originally neutral cues.
Much of the work with this paradigm has been done by
Forsyth and colleagues (Forsyth and Eifert, 1998; Forsyth et al.,
1996) who studied traditional measures of fear conditioning,
such as electrodermal responses, heart rate, subjective units of
distress, and panic symptoms. In one of those studies (Forsyth
and Eifert, 1998), video fragments varying in fear-relevance
(snake, heart beating, and flowers) were paired with 20 s
inhalations of 20% or 13% CO2-enriched air. Evidence for
stronger fear conditioning to the fear-relevant compared to the
fear-irrelevant video fragments was found, both in the
autonomic indices and the subjective reports.
Former studies from our laboratory typically applied odors
or mental imagery as CSs and 2 min inhalations of 7.5% or
5.5% CO2-enriched air as the US in a differential conditioning
paradigm (for reviews, see: Van den Bergh et al., 2001, 2002).
Results showed that participants easily learn to report bodily
symptoms in response to an unpleasant CS+ odor after only
three pairings of the respective odor with the CO2-inhalation.
Importantly, these effects tended to be more pronounced in
participants scoring high on Negative Affectivity (NA)
Author's personal copy
88
S. Fannes et al. / Biological Psychology 78 (2008) 87–92
(Devriese et al., 2000; Van den Bergh et al., 1998) and in
psychosomatic patients (Van den Bergh et al., 1997). In some
studies, also breathing behavior was conditioned, but the
conditioning effect on respiration lacked consistency across
studies (Van den Bergh et al., 1995, 1997). A first reason for this
may be that relatively broad time windows (2 min) were used
for the averaging of the breathing parameters, potentially
masking transient, short-lived conditioned changes in breathing
behavior. In addition, participants may differ in the way they
show primarily instrumental or respondent conditioning of their
breathing behavior. Whereas some may try to actively inhibit
inhaling the CO2 (leading to a conditioned instrumental
decrease in breathing), others may show a respiratory CR
similar to the unconditional effects of CO2 (increased
breathing). It can be expected that the avoidant pattern
(decreased ventilation) is most prominent at the onset of a CO2inhalation, whereas the unconditional effects of CO2 (increased
ventilation) are more likely to occur towards the end of a trial.
Attempts to avoid CO2-inhalation have been described both in
humans (Lejeuz et al., 1998) and animals (Raj and Gregory,
1995). Nonetheless, most human fear conditioning studies using
CO2-inhalation have focused on respondent classical conditioning and have overlooked instrumental avoidance of the CO2 as a
potential behavioral index of fear. Particularly during inhalations
of short durations (e.g., 20 s), trying not to inhale the CO2 is likely
an active way to cope with the aversive event.
A first aim of the current study was to provide a detailed
description of changes in ventilation and fractional end-tidal
CO2 across time in a differential fear conditioning paradigm
pairing an odor (CS) with CO2-inhalation (US). A relative
strong US (20% during 30 s) was opted for to maximize the
chance to observe fear conditioning. We expected that
avoidance in breathing behavior as indicated by a decrease
in minute ventilation would develop across acquisition trials in
response to the odor paired with the CO2 (CS+), but not in
response to the control odor (CS ).
A second aim was to explore the relation of avoidance
behavior in respiration with subjective anxiety and USexpectancy. We expected more subjective anxiety and higher
US-expectancy ratings in participants showing conditioned
avoidance in breathing behavior compared to participants not
avoiding the CO2.
1. Method
1.1. Participants
Fifty-eight healthy participants (29 men and 29 women, 55 undergraduate
students/3 community members, 53 Caucasian/5 Asian, mean age 22, age range
18–55 years) volunteered in return for course credit or 10s. Participants were
only allowed to participate if they confirmed not to (have) suffer(ed) from any
major respiratory or cardiac disease, epilepsy, or psychiatric disorder. Data from
three participants were excluded from analyses because of technical problems.
1.2. Materials
1.2.1. Subjective measures
Participants completed the Checklist of Psychosomatic Symptoms (CPS,
Wientjes and Grossman, 1994) measuring the occurrence of 35 symptoms in daily
life, the Anxiety Sensitivity Index (Reiss et al., 1986) and the Dutch version of the
Positive and Negative Affect Schedule (Engelen et al., 2006) before the start of the
experiment. Following each breathing trial, they completed a state version of the
CPS, but the latter results are beyond the scope of this paper and will not be
discussed. Upon presentation of the CS, participants rated their online anxiety
(‘‘How anxious are you now?’’) and US-expectancy (‘‘To what extent do you
expect to experience bodily complaints in the present trial?’’) on a 7-point bipolar
scale [ranging from 3 (not at all) to +3 (very much)].
1.2.2. Apparatus and software
Participants wore a CO2 nasal sampling cannula and a face mask (8900
Series, Hans RudolphTM) connected to a flow meter (Fleish no. 2, Epalinges,
Switzerland). Upstream from the latter device, a non-rebreathing valve ensured
the separation of inspiratory and expiratory air. A vinyl tube (inner diameter:
3.5 cm; length 100 cm) connected the inspiratory side of the non-rebreathing
valve with a three-way Y-valve (stopcock type). The latter enabled easy
switching between room air and air from a meteorological balloon containing
a decompressed mixture of 20% CO2, 17% O2, and 63% N2.
The odors were being vaporized using a DevilBiss 646 nebulizer at a
constant airflow of 2 L/min. Small vinyl tubes connected the nebulizer to the
side of the mask, allowing the mixing of the odor with the inspiratory gas. Two
foul-smelling odors were used: diluted ammonia (0.8%) and butyric acid
(100%).
The signals from the infrared CO2-monitor (Poet II, Criticare, Waukesha,
WI), and the pressure transducer (Sine Wave Carier Demodulator CD15,
Valydine EngineeringTM) were sampled at 20 Hz and were daily calibrated
using a 7.5% CO2 mixture and a 1 L syringe, respectively.
Both the CO2 and the flow signal were treated off-line with PSPHA (De
Clerck et al., 2006), a modular script-based program which we further developed to generate and apply calibration factors for each signal and to extract the
following parameters for each breath: end-tidal CO2-pressure (FetCO2, in %)
and minute ventilation (in ml/min). All waveforms were visually inspected offline and technical abnormalities and movement artifacts were eliminated using
the PSPHA software.
1.3. Procedure
Participants first received written information about the purpose and
possible adverse effects of the experimental manipulation, then signed the
informed consent form and completed the questionnaires. They were told (a)
that the study was designed to monitor breathing behavior during the inhalation
of several odorous gases; (b) that two innocuous mixtures would be administered, and that one of them could temporarily cause harmless symptoms, such as
shortness of breath, a little dizziness and headache which would disappear
quickly after the trial, while the other mixture would not cause such symptoms;
and (c) that they were allowed to stop the experiment at any time.
All participants started with a context exposure trial (breathing room air
through the system for 2 min in absence of any odor) to get habituated to the
breathing circuit. The acquisition phase consisted of six semi-randomized trials:
three CS+ trials (odor presented together with CO2-enriched air) and three CS
trials (odor presented together with room air). Half of the participants received
ammonia as the CS+ and butyric acid as the CS , whereas this was reversed for
the other half. Breathing trials lasted for 40 s. The US (CO2-enriched air) was
presented 10 s after CS+ onset. Both co-terminated 30 s later. After the
administration of the CS+/US compound, participants continued to breathe
through the mask for another 10 s, to assure that all CO2 was being eliminated
from the tubing system after CS+ trials. During the CS trials, regular room air
was administered instead of CO2-enriched air. Intertrial intervals lasted 4 min
after CS+ trials and 2 min after CS trials. A pause of 5 min was inserted
between acquisition and extinction.
The extinction phase was identical to the acquisition phase, with the
exception that (a) no CO2-enriched air was used in any trial (CS+ only and
CS only trials); and (b) all intertrial intervals lasted 2 min.
Participants were seated in a small room next to the experimenter’s room
and were unable to see the apparatuses. The experimenter gave instructions
through a microphone, manipulated the switches and carefully watched the
participant on a monitor to ensure that the mask remained in place during the
breathing trials.
Author's personal copy
S. Fannes et al. / Biological Psychology 78 (2008) 87–92
1.4. Data analysis and design
89
acquisition, participants learned to decrease their ventilation in
response to the CS+ odor signaling the impending US: a
significant decrease in ventilation from the pre-US (time window
0–10 s, in which only the CS+ odor was present) to the first 10 s
of the CO2-inhalation (time window 10–20 s) was observed in the
first [F(1, 55) = 7.96, p < .01], but not in the second [F(1,
55) < 1)] or the third [F(1, 55) < 1] acquisition trial. No such
pattern was observed for CS trials. Furthermore, the decrease in
ventilation during the first 20 s of CO2 inhalation (time window
10–30 s) tended to be stronger in the first compared to the second
and third acquisition trial [F(1, 55) = 3.54, p = .06].
Averages of FetCO2 and minute ventilation were calculated for each 10 s
interval. Each breath was weighted for its relative time occupied in the 10 s
interval.
In a first set of analyses, minute ventilation and FetCO2 of acquisition and of
extinction were analyzed separately in a CS+odor (butyric acid/
ammonia) CS (+/ ) Trial (1/2/3) Time (4 subsequent 10 s time windows) repeated measures ANOVA. Only CS+odor was a between subject
variable.
A second series of analyses focused on subjective anxiety and US-expectancy as a function of conditioned avoidance in breathing behavior in the last
acquisition trials. To this end, the relative change in minute ventilation during
the last acquisition CS+ trial compared to the last acquisition CS trial was
calculated for each participant. A median split on this index of conditioned
avoidance in breathing behavior during late acquisition was performed to divide
the participants in a subgroup who showed conditioned breathing inhibition
(‘Avoiders’) or not (‘Non-avoiders’). Online subjective anxiety and US-expectancy were analyzed in a repeated measures ANOVA in a group (Avoiders/Nonavoiders) CS (+/ ) Trial (+/ ) design.
Only significant effects will be reported. Greenhouse–Geisser corrections
were applied for all effects involving the trial variable. Uncorrected degrees of
freedom and p’s will be reported together with e.
2.2. Extinction
2.2.1. FetCO2
A significant CS Time Trial interaction was found
[F(6, 318) = 2.33, p < .05, e = .73, h2 = .04, see Fig. 2].
Follow-up comparisons indicated that FetCO2 in the first time
window (0–10) was higher during CS+ than during CS trials
in the first two extinction trials [F(1, 53) = 4.25, p < .05], but
not in the last extinction trial [F(1, 53) < 1]. Additional simple
main tests of this three-way interaction revealed that FetCO2
during the first time window (0–10 s) tended to decrease across
CS+ extinction trials [F(1, 53) = 3.75, p < .06, e = .81], but not
across CS trials [F(1, 53) < 1].
2. Breathing behavior
2.1. Acquisition
2.1.1. FetCO2
A significant CS Time interaction indicated that FetCO2
increased strongly during CS+ trials, but not during CS trials
[F(3, 159) = 311,92, p < .01, e = .74, h2 = .85; see Table 1].
2.2.2. Minute ventilation
Follow-up comparisons of a marginally significant three-way
interaction (CS Time Trial: F(6, 324) = 2.06, p < .06,
e = .82; h2 = .04, see last three panels of Fig. 1) showed that a
linear increasing trend of minute ventilation during the first 10 s
was present across CS+ extinction trials [F(1, 55) = 4.52,
p < .05, e = .82]. This was not observed for CS extinction trials
[F(1, 55) = 0.14, n.s., e = .82]. In other words, the inhibited
ventilation in response to the CS+odor learned during the
acquisition phase, disappeared across extinction.
2.1.2. Minute ventilation
Follow-up comparisons of a significant CS Time interaction [F(3, 165) = 5.71, p < .0001, e = .80, h2 = .09; see
Table 1] indicated a specific response pattern across time during
CS+ trials. Participants’ minute ventilation was lower during
the first 20 s of CO2 inhalation (time window 10–30 s)
compared to the corresponding time window of CS trials
[F(1, 55) = 11.44, p < .01]. Minute ventilation subsequently
increased towards the last 10 s (30–40 s) of CS+ acquisition
trials [F(1, 55) = 24.35, p < .01], but not of CS trials [F(1,
55) = 0.58, n.s.].
However, this response pattern changed across acquisition
trials, as indicated by a significant CS Time Trial interaction [F(6, 330) = 2.59, p < .05, e = .78, h2 = .05 see first three
panels of Fig. 1]. Follow-up comparisons showed that across
3. Self-reported anxiety and US-expectancy
3.1. Avoiders versus Non-avoiders
Descriptive statistics of both groups can be found in Table 2.
The group who avoided the US by inhibiting their ventilation
during the last CS+ compared to the last CS acquisition trial,
scored significantly higher on trait Negative Affectivity [F(1,
Table 1
Means (S.D.s) FetCO2 (in %) and minute ventilation (ml/min) during acquisition and extinction, per CS type and time window
Time window
0–10 s
CS+
FetCO2
Acq
Ext
5.29 (0.47)
5.16 (0.47)
Minute ventilation Acq 7467 (3016)
Ext 8146 (3413)
Note: Acq = acquisition; Ext = extinction.
10–20 s
CS
CS+
5.31 (0.48)
5.13 (0.48)
7608 (2862)
8040 (3492)
6.04 (0.79)
5.24 (0.52)
7060 (2821)
7701 (3065)
20–30 s
CS
CS+
5.38 (0.50)
5.21 (0.54)
7812 (2870)
7480 (2998)
8.28 (1.28)
5.35 (0.53)
6943 (2509)
8097 (2806)
30–40 s
CS
CS+
5.40 (0.51)
5.36 (0.54)
8180 (2953)
7996 (2728)
9.28 (0.96)
5.33 (0.55)
8441 (3911)
7864 (2210)
CS
5.47 (0.50)
5.31 (0.58)
8321 (2761)
8049 (2426)
Author's personal copy
90
S. Fannes et al. / Biological Psychology 78 (2008) 87–92
Fig. 1. Minute ventilation (averaged across 10 s windows) during CS+ and CS
acquisition and extinction trials.
Fig. 2. Fractional end-tidal carbon dioxide (averaged across 10 s windows) during CS+ and CS
extinction trials.
Author's personal copy
S. Fannes et al. / Biological Psychology 78 (2008) 87–92
Table 2
Mean (S.D.s) of Avoiders and Non-Avoiders on relative conditioned change in
minute ventilation during late acquisition, trait Negative Affectivity and
Anxiety Sensitivity scores and total score on the Checklist for Psychosomatic
Symptoms
Avoiders
(N = 29)
%Change MV
NA
ASI
CPS
37% (19%)
20.0 (5.2)
22.10 (6.36)
78.6 (16.3)
Non-avoiders
(N = 28)
27% (55%)
17.3 (4.2)
21.86 (9.26)
73.0 (16.3)
t(54)
5.86**
2.14*
.12
1.29
Note: %change MV = relative increase/decrease in minute ventilation during
the last acquisition CS+ trial compared to the last CS trial (time window 10–
30 s); NA = score on the NA-scale of the PANAS; ASI = Anxiety Sensitivity
Index; CPS: score on the Checklist for Psychosomatic Symptoms.
*
p < .05.
**
p < .01.
54) = 4.59, p < .05, h2 = .08] than the group who did not show
such a conditioned inhibition of breathing.
The ratio of the number of participants who had received
butyric acid versus ammonia as the CS+ was 16/12 for the
Avoiders and 11/16 for the Non-avoiders. Both genders were
equally represented across both groups (Avoiders: 13 women/
15 men; Non-avoiders: 15 women/12 men)
3.2. US-expectancy
No significant effects were observed on US-expectancy
during acquisition. However, a significant group CS interaction showed up during extinction [F(1, 54) = 13.31, p < .01,
h2 = .20]: Whereas a differential conditioning effect on USexpectancy was absent for the Non-avoiders [F(1, 54) = 0.19,
n.s., M = 0.06, S.D. = 1.20, and M = 0.17, S.D. = 1.28, for
CS+ and CS , respectively], the Avoiders had a significantly
higher US-expectancy during CS+ (M = 0.91, S.D. = 0.98),
than during CS (M = 0.48, S.D. = 1.59) extinction trials
[F(1, 54) = 31.28, p < .01].
3.3. Anxiety
A main effect of CS was the only significant effect observed
for the acquisition data [F(1, 54) = 4.17, p < .05, h2 = .07]:
participants reported more anxiety during CS+ (M = 0.46,
S.D. = 1.22) than during CS acquisition trials (M = 0.71,
S.D. = 1.34). Three main effects were observed for the
extinction data. First, participants still felt more anxious
during CS+ (M = 0.82, S.D. = 1.36) than during CS
(M = 1.05, S.D. = 1.36) acquisition trials [main effect of
CS: F(1, 54) = 4.34, p < .05, h2 = .07]. Second, self-reported
anxiety decreased across extinction in a non-differential way
[main effect of Trial: F(2, 108) = 3.25, p < .05, e = .84,
h2 = .06]. Means for Trials 1 through 3 were 0.84, 0.88, and
1.08 (S.D.s = 1.38, 1.35, 1.37). Third, Avoiders reported
overall more anxiety throughout extinction than Non-avoiders
[main effect of group: F(1, 54) = 4.04, p < .05, h2 = .07;
Avoiders: M = 0.60, S.D. = 1.16; Non-avoiders: M = 1.27,
S.D. = 1.35]. No significant interaction effects were present.
91
4. Discussion
The present study examined breathing behavior, fractional
end-tidal CO2, subjective fear, and US-expectancy of healthy
participants in a differential fear conditioning paradigm with
odors as CSs and 20% CO2-inhalation as US. Inhalation of
CO2-enriched air increases arterial CO2-pressure, reflexively
leading to an increase in ventilation. However, no such increase
in ventilation was observed during the acquisition phase of the
present study. In line with our hypothesis, the participants
paradoxically decreased their ventilation during CO2-administrations, indicating active avoidance of the CO2-inhalation.
Importantly, the avoidant breathing pattern changed across
CS+ acquisition trials, suggesting learning processes. Whereas
participants showed a decrease in ventilation from the pre-US
time window (CS+ odor only, 0–10 s) to the first US time
window (CS+ odor and CO2) in the first acquisition trial, this
pattern was no longer observed in subsequent acquisition trials.
From the second acquisition trial on, participants already
decreased their ventilation in response to the CS+ odor before
they were switched to the CO2-breathing (0–10 s time window).
This effect disappeared during subsequent extinction. In other
words, breathing inhibition to the CS+odor was established in
acquisition and subsequently extinguished.
The finding that participants actively tried not to inhale the
CO2 indicates the importance of studying the temporal
dynamics of respiratory behavior in fear paradigms applying
CO2-inhalation, especially if one wants to investigate the
conditioning of respiratory behavior. Indeed, no consistent
conditioning effects on respiration may have been found in
former studies because of the assumption that CO2-inhalation
induces a reflexive, linear increase in ventilation and arousal.
Besides these temporal dynamics, the present study has also
documented substantial interindividual variability in the breathing response to CO2. When participants were split on a post-hoc
basis into a group showing conditioned avoidance in breathing
behavior (‘Avoiders’) or not (‘Non-avoiders’), it appeared that
the Avoiders scored higher on Negative Affectivity than the Nonavoiders and only the Avoiders showed a higher US-expectancy
to the CS+ compared to the CS . Interestingly, subjective
anxiety to the CSs did not change from acquisition to extinction
for the Avoiders, whereas it decreased for the Non-avoiders.
These results strongly suggest that avoidance of the US may have
implications for resistance to extinction. Indeed, disconfirmation
of the US-expectancy during extinction may be much harder to
perceive for people who did not fully confront the US during the
preceding acquisition phase, resulting in continued high levels of
anxiety throughout extinction. More generally, avoidance CO2inhalation may constitute a behavioral index of fear that may be
particularly relevant in the context of disorders in which
avoidance of interoceptive sensations of arousal may be
apparent, e.g., panic disorder or cardiophobia. Particularly, the
paradigm and measures of the present study may allow to study
the dynamics of how the (Non-)avoidance of interoceptive
sensations of arousal may influence subsequent fear and anxiety.
Some limitations of the present study should be acknowledged and a few directions for future research can be outlined.
Author's personal copy
92
S. Fannes et al. / Biological Psychology 78 (2008) 87–92
First, given the explorative nature of the present study, the results
clearly need replication. In addition, future studies may also
investigate the role of interindividual difference variables, such
as Fear of Suffocation, Anxiety Sensitivity or Avoidant Coping
Style (Zvolensky et al., 2001; Spira et al., 2004) as a predictor of
active avoidance in breathing behavior and/or resistance to
extinction of fear. A second concern is the possibility that
participants may have tasted/sensed the CO2 from the very first
breath onwards, that is, before the CO2 could reflexively have
increased autonomic arousal. Given the dead space of the
breathing circuitry connecting the face mask with the balloon
containing CO2-enriched air, it takes a few breaths before the
CO2-enriched air reaches the participant and can start acting
upon the chemoreceptors. Therefore, it is somehow puzzling to
observe that our participants already showed a decrease in
ventilation in the first 10 s following the switching to the CO2
(time window 10–20 s, see Fig. 1). One interpretation may be that
the participants have tasted/sensed the CO2 from the very first
breath onwards. Several findings are consistent with this
interpretation. First, no such decrease in ventilation in the
second time window (10–20 s) was observed in the first
extinction trial, which is consistent with the idea that the taste
may have been a predictor of the US, in addition to the CS+ odor.
During extinction, participants may not have reduced their
ventilation during that time window anymore, because there was
no taste of the CO2. Second, the possibility of the taste of CO2
functioning as a CS has been mentioned earlier (Gallego and
Perruchet, 1991). From a conditioning perspective, an additional
taste-CS may, in part, have overshadowed the odor-CS,
compromising conditioning to some extent. Future studies
should clarify whether the observed avoidance in breathing
behavior is dependent on a potential taste-experience upon the
first breath containing CO2-enriched air and how this ‘unwanted’
taste-CS can be controlled for.
In sum, the present findings clearly point to the importance
of carefully measuring respiratory behavior in fear paradigms
applying CO2-challenges. On the one hand, it provides
important information about the respiratory effects of the
US. On the other hand, it is necessary to detect learned
avoidance in breathing behavior. As such, respiratory response
patterns may be used as a behavioral index of fear,
complementary to the more traditionally used measures of
self-reported fear, electrodermal activity and heart rate.
Acknowledgement
I. Van Diest and D. Vansteenwegen are post-doctoral fellows
of the Fund of Scientific Research – Flanders (FWO). This
study was financed by a grant of the FWO (G.0553.05).
References
Acheson, D.T., Forsyth, J.P., Prenoveau, J.M., Bouton, M.E., 2007. Interoceptive fear conditioning as a learning model of panic disorder: an experimental
evaluation using 20% CO2-enriched air in a non-clinical sample. Behaviour
Research and Therapy 45, 2280–2294.
Barlow, D.H., 1988. Anxiety and its Disorders: the Nature and Treatment of
Anxiety and Panic. Guilford Press, New York.
De Clerck, A., Verschuere, B., Crombez, G., De Vlieger, P., 2006. Psychophysiological analysis (PSPHA): a modular script based program for
analyzing psychophysiological data. Behavior Research Methods 38 (3),
504–510.
Devriese, S., Winters, W., Stegen, K., Van Diest, I., Veulemans, H., Nemery, B.,
Eelen, P., Van de Woestijne, K., Van den Bergh, O., 2000. Generalization of
acquired somatic symptoms in response to odors: a Pavlovian perspective on
multiple chemical sensitivity. Psychosomatic Medicine 62, 751–759.
Devriese, S., De Peuter, S., Vos, G., Van Diest, I., Van de Woestijne, K.P., Van
den Bergh, O., 2006. US-inflation in a differential odour conditioning
paradigm is not a robust phenomenon. Journal of Behavior Therapy and
Experimental Psychiatry 37, 314–332.
Engelen, U., De Peuter, S., Victoir, A., Van Diest, I., Van den Bergh, O., 2006.
Verdere validering van de ‘‘Positive and Negative Affect Schedule’’
(PANAS): twee nederlandstalige versies vergeleken. Gedrag & Gezondheid
34 (2), 89–102.
Forsyth, J.P., Eifert, G.H., 1996. Systemic alarms in fear conditioning I: a
reappraisal of what is being conditioned. Behavior Therapy 27, 441–462.
Forsyth, J.P., Eifert, G.H., 1998. Response intensity in content-specific fear
conditioning comparing 20% versus 13% CO2-enriched air as unconditioned stimuli. Journal of Abnormal Psychology 107 (2), 291–304.
Forsyth, J.P., Eifert, G.H., Thompson, R.N., 1996. Systemic alarms and fear
conditioning. II. An experimental methodology using 20% carbon dioxide
as an unconditioned stimulus. Behavior Therapy 27, 391–415.
Gallego, J., Perruchet, P., 1991. Classical conditioning of ventilatory responses
in humans. Journal of Applied Physiology 70 (2), 676–682.
Griez, E., Colasanti, A., van Diest, R., Salamon, E., Schruers, K., 2007. Carbon
dioxide inhalation induces dose-dependent and age-related Negative Affectivity. Plos ONE 2 (10), e987 doi:10.1371/journal.pone.0000987.
Lejeuz, C.W., O’Donnell, J., Wirth, O., Zvolensky, M.J., Eifert, G.H., 1998.
Avoidance of 20% carbon dioxide-enriched air with humans. Journal of the
Experimental Analysis of Behavior 70, 79–86.
Raj, A.B.M., Gregory, N.G., 1995. Welfare implications of the gas stunning of
pigs 1. Determination of aversion to the initial inhalation of carbon dioxide
or argon. Animal Welfare 4, 273–280.
Reiss, S., Peterson, R.A., Gursky, D.M., McNally, R.J., 1986. Anxiety sensitivity, anxiety frequency, and the prediction of fearfulness. Behaviour
Research and Therapy 24, 1–8.
Spira, A.P., Zvolensky, M.J., Eifert, G.H., Feldner, M.T., 2004. Avoidanceoriented coping as a predictor of panic-related distress: a test using
biological challenge. Anxiety Disorders 18, 309–323.
Van den Bergh, O., Kempynck, P.J., Van de Woestijne, K.P., Baeyens, F., Eelen,
P., 1995. Respiratory learning and somatic complaints: a conditioning
approach using CO2-inhalation. Behaviour Research and Therapy 5,
517–527.
Van den Bergh, O., Stegen, K., Van de Woestijne, K.P., 1997. Learning to have
psychosomatic complaints: conditioning of respiratory behavior and
somatic complaints in psychosomatic patients. Psychosomatic Medicine
59, 13–23.
Van den Bergh, O., Stegen, K., Van de Woestijne, K.P., 1998. Memory effects on
symptom reporting in a respiratory learning paradigm. Health Psychology
17 (3), 241–248.
Van den Bergh, O., Devriese, S., Winters, W., Eelen, P., Veulemans, H., Nemery,
B., Van de Woestijne, K.P., 2001. Acquiring symptoms in response to odors:
a learning perspective on multiple chemical sensitivity. In: Sorg, B., Bell, I.
(Eds.), The Role of Neural Plasticity in Chemical Intolerance, Annals of
the New York Academy of Sciences 933, 278–290.
Van den Bergh, O., Winters, W., Devriese, S., Van Diest, I., 2002. Learning
subjective health complaints. Scandinavian Journal of Psychology 43 (2),
147–152.
Wientjes, C.J.E., Grossman, P., 1994. Overreactivity of the psyche or the soma?
Interindividual associations between psychosomatic symptoms, anxiety,
heart rate, and end-tidal partial carbon dioxide pressure. Psychosomatic
Medicine 56, 533–540.
Zvolensky, M.J., Feldner, M.T., Eifert, G.H., Stewart, S.H., 2001. Evaluating
differential predictions of emotional reactivity during repeated 20%
carbon dioxide-enriched air challenge. Cognition and Emotion 15 (6),
767–786.