Psychiatry Research 209 (2013) 566–573

Contents lists available at ScienceDirect

Psychiatry Research journal homepage: www.elsevier.com/locate/psychres

Experiential, autonomic, and respiratory correlates of CO2 reactivity in individuals with high and low anxiety sensitivity Jens Blechert a,n, Frank H. Wilhelm a, Alicia E. Meuret b, Eva M. Wilhelm c,d, Walton T. Roth c,d a

Department of Psychology, Division of Clinical Psychology, Psychotherapy, and Health Psychology, University of Salzburg, Salzburg, Austria Department of Psychology, Southern Methodist University, Dallas, TX, USA c Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA, USA d Department of Veterans Affairs Health Care System, Palo Alto, CA, USA b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 June 2012 Received in revised form 26 November 2012 Accepted 3 February 2013

Psychometric studies indicate that anxiety sensitivity (AS) is a risk factor for anxiety disorders such as panic disorder (PD). To better understand the psychophysiological basis of AS and its relation to clinical anxiety, we examined whether high-AS individuals show similarly elevated reactivity to inhalations of carbon dioxide (CO2) as previously reported for PD and social phobia in this task. Healthy individuals with high and low AS were exposed to eight standardized inhalations of 20% CO2-enriched air, preceded and followed by inhalations of room air. Anxiety and dyspnea, in addition to autonomic and respiratory responses were measured every 15 s. Throughout the task, high AS participants showed a respiratory pattern of faster, shallower breathing and reduced inhalation of CO2 indicative of anticipatory or contextual anxiety. In addition, they showed elevated dyspnea responses to the second set of air inhalations accompanied by elevated heart rate, which could be due to sensitization or conditioning. Respiratory abnormalities seem to be common to high AS individuals and PD patients when considering previous findings with this task. Similarly, sensitization or conditioning of anxious and dyspneic symptoms might be common to high AS and clinical anxiety. Respiratory conditionability deserves greater attention in anxiety disorder research. & 2013 Elsevier Ireland Ltd. All rights reserved.

Keywords: CO2 reactivity Heart rate Respiration Anxiety sensitivity Panic disorder Tidal volume Classical conditioning

1. Introduction Anxiety sensitivity (AS) refers to a persistent fear of anxietyrelated bodily sensations based on the belief that these symptoms have harmful physical, psychological, or social consequences (Reiss, 1991; Smits et al., 2008). Over the last three decades, this personality attribute has given rise to many investigations and is a current focus of considerable attention (reviews in Olatunji and Wolitzky-Taylor, 2009; Naragon-Gainey, 2010). Interest in AS stems from its role as predictor of general psychopathology and particularly of PD. Even though the bulk of research has focused on panic disorder (PD), AS has been found to be relevant to a range of mental disorders. AS measured by the Anxiety Sensitivity Index (ASI, Reiss et al., 1986) is associated cross-sectionally with depression (Rodriguez et al., 2004), posttraumatic stress disorder (Lang et al., 2002; Kilic et al., 2008; Marshall et al., 2010), social and generalized anxiety disorders (e.g., Deacon and Abramowitz, 2006), obsessive compulsive disorder (McWilliams et al., 2007), n Correspondence to: Division of Clinical Psychology, Psychotherapy, and Health Psychology, University of Salzburg, Austria, Hellbrunnerstrasse 34, 5020 Salzburg, Austria. Tel.: þ 43 662 8044 5163. E-mail address: [email protected] (J. Blechert).

0165-1781/$ - see front matter & 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.psychres.2013.02.010

and several other disorders (Asmundson and Taylor, 1996; Joiner et al., 2008; Olatunji et al., 2009). Longitudinal studies support the role of AS as a precursor of panic attacks and a diagnosis of PD but also of other anxiety disorders (Hayward et al., 2000; Schmidt et al., 2006). It is currently unclear whether there is a specific association between AS and PD over and above AS being a predictor of anxiety disorders or general psychopathology. AS is a cognitive construct concerning beliefs about the harmfulness of somatic symptoms. Several anxiety disorders and PD in particular are characterized not only by high AS, but also by prominent physiological anxiety symptoms, such as breathlessness, racing heart, or sweating (American Psychiatric Association, 1994). Thus, the question arises whether AS mediates a primarily cognitive risk or whether some psychophysiological vulnerability is present in high AS individuals, too. Thus, research on the relationship between AS and PD should profit from measuring precisely both their psychological and physiological aspects. Provoking anxiety in the laboratory would allow the investigator to study this relationship under defined and controlled conditions. Recent years have seen an increased interest in the psychophysiological and neuronal correlates of AS among a range of different laboratory conditions (e.g., Killgore et al., 2011; Melzig et al., 2011; McMillan et al., 2012) demonstrating a broad

J. Blechert et al. / Psychiatry Research 209 (2013) 566–573

spectrum of altered stress- and anxiety-reactivity in this group. However, given the prominence of respiratory symptoms in PD (Niccolai et al., 2009; Grassi et al., 2013, but see also Pfaltz et al., 2009), particularly in the respiratory subtype of PD (Abrams et al., 2006; Freire et al., 2008, 2010; Kircanski et al., 2011; RobersonNay et al., 2012) one line of psychophysiological research has focused on respiratory symptoms in potential precursors and risk factors for PD such as AS. A prominent respiratory challenge in PD is the inhalation of carbon dioxide-enriched air (referred to here as carbon dioxide (CO2)), which induces a pattern of symptoms (including breathlessness, increased ventilation, and dizziness), that resembles naturally occurring panic attacks to a relatively high degree (Rassovsky and Kushner, 2003). A number of reports have found elevated CO2 reactivity – self-reported panic and anxiety symptoms in response to inhalation of high concentrations of CO2 (sometimes termed CO2 sensitivity, see Vickers, 2012, for issues of definition) – in PD patients relative to patients with other anxiety disorders or healthy individuals (van den Hout et al., 1987; Gorman et al., 1990; Griez et al., 1990; Caldirola et al., 1997; Papp et al., 1997; Perna et al., 2003), and in individuals with higher compared to those with lower AS (e.g., Asmundson et al., 1994). However, the customary protocol that administers single or double breath high concentration CO2 (typically 35%) and focuses on panic rate (number of participants reporting panic in response to the challenge) overlooks less intense arousal/anxiety episodes and associated learning/sensitization processes. Furthermore, the panic naı¨ve, high AS participants may report their anxiety experience differently from people who have had repeated spontaneous panic attacks. More recent laboratory based studies have used medium-dosed CO2 concentrations (e.g., 20%) which allow multiple inhalations and sensitive measurement of multiple autonomic and respiratory variables (Acheson et al., 2007, 2012; Fannes et al., 2008). This allows for a more fine-grained analysis of the temporal dynamics of the evolving physiological and subjective anxiety responses in either anxiety patients or high risk individuals. A series of studies by Beck and colleagues found that a subsample of high AS individuals (Beck et al., 1996, 1999; Beck and Wolf, 2001) and patients with PD (Beck and Shipherd, 1997) exhibited sensitization—increased anxious responding across repeated CO2-inhalations instead of the expected habituation which might explain how repeated panic experiences can lead to PD in these ‘non-habituators’ (see Li et al., 2006, 2008, for a similar line of research in trait anxiety). However, the absence of low AS control groups and of autonomic and respiratory measures limits conclusions about mechanisms common to AS and PD. Telch et al. (2011) manipulated cognitive expectancies about the consequences of a single CO2 inhalation in high- and low-AS individuals and showed that negative expectancies interacted with AS in triggering panic. The role of emotional avoidance and context conditioning was emphasized by Forsyth and coworkers (Finlay and Forsyth, 2009; Kelly and Forsyth, 2009). However, again, respiratory variables were not measured. We have recently used repeated, medium-dosed CO2 inhalations in a task that delivers eight inhalations of 20% CO2 enriched air, preceded and followed by room air inhalations, which allows for an assessment of dependent measures every 15 s. This high temporal resolution allowed us to assess response trajectories on two levels: on a ‘macro-level’ – across inhalations (i.e., habituation, sensitization), and on a ‘micro-level’ – within inhalations (i.e., baseline, peak reactivity, and recovery). Patients with PD, patients with social phobia (SP), and healthy controls underwent this task, giving us an opportunity to measure processes both common to anxiety disorders and specific to PD (Blechert et al., 2010). On the micro-level, both anxious groups showed increased reactivity and delayed recovery of subjective and autonomic responses to the CO2 inhalations, suggesting

567

these two response patterns characterize clinical anxiety in general. Furthermore, both SP and PD individuals manifested continued subjective and physiological responding to the air inhalation trials that followed the CO2 inhalations, indicating that some kind of sensitization or fear learning had taken place. However, on the macro-level, only PD patients gave non-habituating, non-declining respiratory responses (tidal volume) across all eight CO2 inhalations, which is in line with research (Maddock and Carter, 1991; Dager et al., 1995; Wilhelm et al., 2001b; for review, Meuret and Ritz, 2010; Gorman et al., 2004; Blechert et al., 2007a; Meuret et al., 2011) and theory (Ley, 1985; Klein, 1993, 1994) linking respiratory abnormalities with PD. Our previous study elucidated the temporal characteristics of potential etiological factors for anxiety disorders in general – at least in so far as SP and PD are representative – and PD specifically. It is difficult, however, to study etiological factors in already sick individuals. Thus, the present study administered the same task to healthy individuals with high AS, and therefore to a population with a putative risk factor for PD, to determine if they differ from their low AS counterparts, and if they do, whether their response patterns would resemble those in our previous study in PD and SP. If high AS was a general predictor for anxious psychopathology we would expect increased responding and delayed recovery of subjective and psychophysiological responses during CO2 inhalations and the last two air inhalations in high AS individuals compared to low AS controls. If high AS had also a specific association with PD, we would expect respiratory abnormalities such as a non-declining ventilatory response to all eight CO2 inhalations. 2. Methods 2.1. Participants Participants were 53 undergraduates at Stanford University, none of whom had participated in our previous study (Blechert et al., 2010). They were recruited to fulfill their psychology class requirements from a larger sample of 414 who had completed a battery of questionnaires including the Anxiety Sensitivity Index (ASI, Reiss, et al., 1986). In line with recommended cutoff scores (Holloway and McNally, 1987), individuals with ASI total scores equal or below 11 and equal or above 20 were contacted by e-mail. Forty-nine Stanford undergraduates with high (n¼27) and low scores (n ¼22) agreed to participate for USD $50. In the week before testing, participants filled out several questionnaires that had been mailed to them. These included a second administration of the ASI. Participants were categorized on the basis of the average of their two ASIs, the second of which tended to regress towards the mean. An average of 16.5 was used to define the groups (high AS group Z17, low AS group r16). The mean ASI score of the high AS group (M¼25.3, S.D.¼3.58, range 17.4–39.0) was thus much higher than the mean of the low AS group (M¼ 9.11, S.D.¼5.85, range 1.0–15.5, t(49)¼ 11.4, Po0.001, d¼ 3.35). Individuals with cardiovascular diseases, respiratory diseases, epileptic seizures, asthma, thyroid problems, neurological disorders, migraine headaches, pregnancy, or current medication use that could alter their physiological or psychological response during testing were excluded. In addition, students with a current or past Axis-I disorder as determined by the Structured Clinical Interview for DSM-IV Axis 1 Disorders (First et al., 1995, administered by a specifically trained clinical psychologist) were excluded. Participants were instructed to abstain from recreational drugs during the 12 h prior to laboratory testing, not drink alcohol in excess, and not use more caffeine or nicotine than they usually do. 2.2. Procedures The study was approved by the local ethics committee for medical research and participants gave written consent before participating. Testing took place in a sound-attenuated chamber partitioned into testing and experimenter rooms. The experimenter could communicate with participants by intercom and observe them through a one-way mirror. The inhalation task was very similar to the one applied to PD and SP previously (Blechert et al., 2010)1 that comprised 12 single breath inhalations.

1 The present and the previous studies (Blechert et al., 2010) differed slightly. In the present study, there were 4 min recoveries after the inhalation and pre-task quiet sitting baselines. The previous study had 3 min recovery after the inhalation and no pre-task baselines.

568

J. Blechert et al. / Psychiatry Research 209 (2013) 566–573

Inhalations one, two, 11, and 12 were of room air, and three through 10 were of 20% CO2 þ 80% O2. This concentration of CO2 is able to trigger substantial anxiety symptoms, but usually not with an intensity that would lead participants to refuse repeated inhalations (Forsyth and Eifert, 1998). Participants were informed in advance about the order of gases and reminded by the rating log sheet, which indicated which gas would occur on each inhalation. The taste and symptoms produced by inhaling 20% CO2 are unmistakably different from room air, so blinding would have been difficult. Each inhalation followed the same sequence: at the beginning of a 15 s acoustic countdown, the experimenter entered the room, placed a clip on the participant’s nose, and handed him/her a tube that came out of a hole in the partition. Five seconds before the end of the countdown, participants were instructed to exhale completely, and at the end of the countdown to take as deep a single breath as possible from the tube and to hold it for 3 s before exhaling. The experimenter then removed the nose clip, put the tube back in place and left the testing room. After 3 min 45 s, during which 15 ratings of anxiety and dyspnea were made (see below), a 15-s countdown began the next inhalation. Quiet sitting baselines preceded (2 min) and followed (4 min) the set of 12 inhalations. See Fig. 1 for an illustration of the task. 2.3. Measures 2.3.1. Self-report measures The ASI is a 16-item questionnaire with five-point scales from zero (very little) to four (very much). Its Cronbach’s Alpha is about 0.8 and its test–retest correlation after three years is 0.7 (Reiss et al., 1986; Maller and Reiss, 1992). The ASI has been revised (e.g., Taylor et al., 2007) but we retained the original version for reasons of comparability with previous research. Participants also completed the Beck Depression Inventory (BDI; Beck et al., 1961), and the StateTrait Anxiety Inventory (STAI) Trait Form (Spielberger et al., 1970). Furthermore, participants completed a questionnaire that inquired about their weekly physical activity in the last 6 months (Ainsworth et al., 2000). In line with our previous research, anxiety and shortness of breath (termed ‘‘dyspnea’’ in the following) were chosen as main subjective dependent variables (Wilhelm et al., 2001a, 2001b; Burkhardt et al., 2010). Participants rated these two items (first anxiety, second dyspnea) during the procedure on 11 point scales from zero (‘‘not at all’’) to 10 (‘‘extremely’’) by circling the appropriate numbers on a rating sheet every 15 s (signaled on a computer display) throughout the inhalations and pre- and post-quiet sitting baselines (see Vickers, 2012, for a discussion of potential problems with assessing subjective anxiety in non-clinical individuals). The first rating was instructed 15 s after the inhalation. Thus, each 4 min inhalation yielded 15 ratings (no rating during the inhalation breath) and the quiet sitting baselines yielded four (pre-) and eight (post-) ratings.

into four consecutive epochs, each containing four physiological segments and four subjective ratings. The 1-min baseline epoch preceded the inhalation and contained the 15 s countdown. Baseline was followed by the peak epoch containing the gas inhalation,2 which in turn was followed by the recovery1 and recovery2 epochs (see Fig. 1 for epochs). The anxiety and dyspnea ratings during quiet sitting preceding and following the 12 inhalations were collapsed into single pre-/postinhalation anxiety and dyspnea scores. Across all 12 inhalations, preliminary trend analyses revealed no linear or non-linear effects and showed that all group effects could be captured by a simplified analysis averaging across pertinent inhalations: air inhalations one and two were averaged into AIR-A, air inhalations three and four (following CO2 inhalations) into AIR-B, CO2 inhalations 1–4 into CO2-A, and CO2 inhalations 5–8 into (illustrated by brackets in Fig. 1). The two levels of the Time factor during CO2 inhalation (CO2-A, CO2-B) allow for an approximation of habituation or sensitization (e.g., Beck et al., 1999). The resulting means scores were then screened for outliers and non-normality. Extreme outlier values outside of 73 S.D.s from the mean were excluded.3 No significant gender differences were found.

2.4.2. Statistical analysis To protect against cumulative alpha error due to multiple testing, multivariate analyses of variance (MANOVAs) preceded ANOVAs for individual measures. Separate repeated measures MANOVAs tested for overall Group effects in ratings of anxiety and dyspnea and for physiological measures. When significant on the Group factor (Wilks’s criterion), MANOVAs were followed by ANOVAs and t-tests for the individual measures as described below. Air inhalations were analyzed separately from CO2 inhalations in a 2  4  2, Time (AIR-A, AIR-B)  Epoch (baseline, peak, recovery1, recovery2)  Group (high AS, low AS) analysis of variance (ANOVA) with repeated measures on Time and Epoch. CO2 inhalations were analyzed in 2  4  2 ANOVAs, with the factors Time (CO2-A, CO2-B)  Epoch (baseline, peak, recovery1, recovery2)  Group (high AS, low AS). In these analyses, increased reactivity (group differences during peak) and/or reduced recovery (group differences during recovery) would be evident in Group  Epoch interactions, and habituation/sensitization in Group  Time interactions. Generally, for reasons of brevity, only effects involving Group are reported.4 The alpha level was set to 0.05; alpha levels between 0.05 and 0.1 were considered as trends. When the sphericity assumption was violated, the Huynh–Feldt correction for repeated measures was applied. Nominal d.f. values and effect sizes are reported (partial eta-squared in % [Z2] for ANOVAs and Cohen’s d for post-hoc ttests).

3. Results 2.3.2. Physiological measures The following dependent measures were assessed to capture sympathetic and parasympathetic autonomic activation and respiratory functioning using a Vitaport I system (Meditec, Karlsruhe, Germany) at a sample rate of 400 Hz and using customized software (Wilhelm and Peyk, 2005).

1. Interbeat interval (IBI) was calculated in ms between successive R-waves in the lead II electrocardiogram after values from ectopic or other kinds of abnormal beats were deleted and replaced by linearly interpolated values. From the IBI instantaneous heart rate (HR) was calculated. 2. Respiratory sinus arrhythmia (RSA) as a measure of cardiac vagal control was calculated using complex demodulation of IBI time series for the frequency band of 0.15–0.4 Hz (Wilhelm et al., 2005). 3. Two channels of respiration were recorded with inductive plethysmography using Respibands (Respitrace Corporation, Ardsley, NY) placed around the chest and abdomen. Calibration against spirometry was accomplished by the least-squares method. Instantaneous respiratory rate, tidal volume, and minute volume were calculated breath-by-breath using customized programs. 4. Expiratory end-tidal partial pressure of CO2 (pCO2) was measured continuously using a calibrated infrared capnograph (N-1000, Nellcor, Hayward, CA) and a dual nostril prong. Subjects were instructed to breathe only through their nose (as confirmed via surveillance through a one-way mirror). End-tidal pCO2 was determined as the level at which pCO2 stopped rising at the end of an expiration (final maximum). 5. Skin conductance level (SCL) was recorded from the middle phalanges of the index and middle finger of the left hand using 11 mm inner diameter Ag/AgCl electrodes filled with isotonic electrode paste. 2.4. Data reduction and analysis 2.4.1. Data reduction In parallel with subjective ratings, mean scores for physiological channels were extracted for 15-s segments (with the exception of RSA, which was extracted in 1-min segments to increase reliability). One inhalation cycle was broken down

3.1. Psychometrics and sample characteristics As shown in Table 1, both groups were comparable in the proportion of men, years of education, as well as body mass index and physical activity, both of which are known to influence respiratory function (Pelosi et al., 1998). Unfortunately, data on smoking status was not available but smoking is generally rare in this population. The age difference, although significant, was negligible (o1 year). Neither of these variables did significantly influence the pattern of results as determined by preliminary analyses. The groups did not differ on BDI-depression scores but the high AS group scored higher on STAI-Trait. Across groups, ASI was correlated with the STAI-T and the BDI, r(46) ¼0.507, Po0.001 and r(46)¼0.355, P¼0.015, respectively. BDI and STAI were not controlled for in this study since the focus was to characterize AS by recruiting individuals with high vs. low values on the ASI trait measure and not to disentangle contributions of STAI (or BDI) and ASI to CO2 reactivity. Adjustment for a subject characteristic that is correlated with the independent variable of interest using ANCOVA is inadvisable in designs with non-random group assignment (Miller and Chapman, 2001). 2 The peak epoch contained only three segments of physiological data due to unstable physiological data resulting from tube handling and gas inhalation. There was no subjective rating during the first 15 s segment of the peak epoch. 3 One outlier (z43) in SCL and one missing value in end-tidal pCO2 were excluded. 4 Epoch effects were strong on all measures showing that the inhalations reliably induced experiential and physiological changes.

J. Blechert et al. / Psychiatry Research 209 (2013) 566–573

569

Fig. 1. CO2 inhalation task. A pre-inhalation quiet sitting baseline (B), two preceding air inhalations, eight 20% CO2 inhalations, and a second set of two air inhalations (AIR-B), followed by a post-inhalation baseline. During analysis two consecutive air inhalations were averaged, resulting in AIR-A and AIR-B blocks. Similarly, four consecutive CO2 inhalations were averaged, resulting in CO2-A and CO2-B blocks. Each inhalation consisted of baseline (B), inhalation (I), and two recovery periods (R, 1 min each) and yielded 15 ratings of dyspnea and anxiety. Table 1 Means (S.D.) of sample characteristics.

% Men Age (years) Body mass index (kg/m2) STAI Trait Form Beck Depression Inventory Physical activity

Low AS group n¼ 22

High AS group n¼ 27

Statistic

40.9 19.6 7 1.05 22.4 7 2.84 31.8 7 7.10 3.97 7 4.61 11.3 7 6.28

37.0 18.9 7 0.85 22.1 7 3.93 39.6 7 10.6 6.71 7 6.51 13.1 7 6.92

w2(1)¼ 0.077, P ¼ 0.782 t(47)¼2.31, P ¼0.026, d ¼0.33 t(42)¼0.89, P ¼0.377 t(41)¼2.76, P ¼0.009, d¼ 0.86 t(41)¼1.55, P ¼0.130 t(43)¼0.97, P ¼0.369

Note: AS, anxiety sensitivity; ASI, Anxiety Sensitivity Index; STAI, State-Trait Anxiety Inventory.

3.2. Ratings of anxiety and dyspnea The MANOVA of anxiety and dyspnea ratings over all 12 inhalations showed a significant Group effect, F(1, 46)¼4.79, P¼0.013, Z2 ¼17.2%, which was followed up with ANOVAs for individual measures, separately for the inhalations of air and CO2 and for pre- and post-inhalation baselines.

3.2.1. Anxiety and dyspnea ratings during pre- and post-inhalation baselines Fig. 2a and b presents anxiety and dyspnea ratings during the pre- and post-inhalation baselines, the air inhalations (AIR-A, AIRB), and the CO2-inhalations (CO2-A, CO2-B). Anxiety ratings during the quiet sitting baselines were analyzed in a 2  2 Time (pre- vs. post-inhalation)  Group ANOVA. A significant Group effect was due to higher anxiety ratings in the high AS group both before and after the inhalation task, F(1, 47) ¼6.14, P¼0.017, Z2 ¼11.6%. No effects involving Group reached significance for dyspnea ratings.

3.2.2. Anxiety and dyspnea ratings during air inhalations Anxiety ratings during air inhalations were analyzed by a 2  4  2 Time (AIR-A, AIR-B)  Epoch (baseline, peak, recovery1, recovery2)  Group (low AS, high AS) ANOVA. The main effect for Group was significant, indicating more anxiety in high AS participants, F(1, 47)¼10.4, P¼0.002, Z2 ¼18.1%. No other effects involving Group reached significance. The ANOVA of dyspnea ratings during air inhalations showed a three-way interaction, F(3, 141) ¼3.57, P¼0.033, Z2 ¼7.05%, in addition to a Group  Epoch interaction, F(3, 141) ¼4.50, P¼0.023, Z2 ¼8.75%. Inspection of the means in Fig. 2b suggests that high AS participants showed increased reactivity on dyspnea ratings particularly during the peak epochs at AIR-B, a pattern that was absent for AIR-A. Follow-up t-tests confirmed that while peak dyspnea ratings did not differ between groups before CO2 inhalations, t(47) ¼1.52, P¼0.135, d¼0.44, high AS participants gave higher peak dyspnea ratings after CO2 inhalations,

t(47) ¼3.01, P¼0.003, d¼0.89. No group differences were evident during the baseline or recovery epochs, ts o1.64, Ps40.10

3.2.3. Anxiety and dyspnea ratings during CO2 inhalations The 2  4  2 Time (CO2-A, CO2-B)  Epoch (baseline, peak, recovery1, recovery2)  Group (low AS, high AS) ANOVA for anxiety ratings during CO2 inhalations showed a significant Group effect, with high AS participants giving higher anxiety ratings than low AS participants, F(1, 47) ¼8.23, P¼0.006, Z2 ¼11.6%, but no other interactions with Group. No significant Group effects emerged for dyspnea ratings during CO2 inhalations.

3.3. Psychophysiological responses Physiological responses are displayed in Figs. 2 and 3. The MANOVA of the seven physiological measures during all 12 inhalations showed a significant Group effect, F(7, 38) ¼2.62, P¼0.017, Z2 ¼42.8%, which was followed by ANOVAS for the individual measures.

3.3.1. Physiological responses during pre- and post-inhalation baselines The 2  2 ANOVA of the pre- and post-inhalation baselines showed an increase in HR in the high AS group from the preinhalation baseline to the post-inhalation baseline that was not present in the low AS group, Group  Time interaction: F(1, 47)¼7.42, P¼0.009, Z2 ¼13.6% (Fig. 2c). This was confirmed in within-group t-tests separately for each group: while HR increased significantly in the high AS group, t(26)¼ 3.07, P¼0.005, d ¼0.33, it remained low in the low AS group, P¼0.54. The 2  2 ANOVA of tidal volume revealed a trend level Group effect, F(1, 47)¼3.24, P ¼0.078, Z2 ¼6.40, due to slightly lower values in the high AS group (Fig. 3c). There were no group differences in the other physiological measures (Fig. 3).

570

J. Blechert et al. / Psychiatry Research 209 (2013) 566–573

4 Dyspnea (0-10)

Anxiety (0-10)

4 3 2 1

75 72 69 66 63

2 1 0

Respiratory rate (br/min)

Heart rate (beats/min)

0

high AS low AS

3

PRE

AIR-A

CO2-A

CO2-B

AIR-B POST

16 14 12 10

PRE

AIR-A

CO2-A

CO2-B

AIR-B POST

Fig. 2. Means and standard errors for self-reported and psychophysiological responses to air and CO2 inhalations for participants with high vs. low anxiety sensitivity. Panels depict self-reported anxiety (a) and dyspnea (b), heart rate (c), and respiratory rate (d). Note: pre ¼ baseline before inhalation phase; AIR-A ¼first set of two inhalations of air; CO2-A¼ first set of four inhalations of CO2; CO2-B ¼second set of four inhalations of CO2; AIR-B ¼second set of two inhalations of air. See Fig. 1 and text for details. Please note that respiration rate drops markedly during the peak of inhalation of air because a full-capacity breath leads to a sudden decline in arterial pCO2. This is being sensed quickly by arterial chemoreceptors and respiratory rate is lowered to raise pCO2 back to normal levels. During a full-capacity breath of 20% CO2 this lowering of arterial pCO2 does not occur (see also Fig. 3, panel b).

3.3.2. Physiological responses during air inhalations The 2  4  2 ANOVA of HR during air inhalations showed a significant Group  Time interaction, F(1, 47)¼12.6, Po0.001, Z2 ¼21.2%. Post-hoc between-group t-tests for each epoch indicated that groups did not differ during any epoch at AIR-A, Ps40.30, but high AS participants had higher HRs during peak and recovery1, Ps o0.048, ds40.59 at AIR-B (Ps for baseline/ recovery2 ¼ 0.08/0.11, respectively). Analysis of respiratory rate during air inhalations showed an Epoch  Group interaction, F(3, 141) ¼3.37, P¼0.046, Z2 ¼ 6.7% (Fig. 2d). Follow-up ANOVAs breaking down the Epoch factor indicated that groups did not differ during peak respiration rate, P40.8, but that high AS participants had higher respiration rates during baseline, F(1, 47) ¼4.06, P ¼0.049, Z2 ¼8.0%, and to a lesser extent during the recovery2, F(1, 47)¼ 3.57, P¼ 0.065 than did low AS participants. Analysis of tidal volume again showed a trend level Group effect, F(1, 47)¼3.82, P¼0.057, Z2 ¼7.53%, due to slightly lower values in the high AS group. There were no group differences on the other physiological measures (RSA, tidal volume, minute ventilation, SCL).

3.3.3. Physiological responses during CO2 inhalations For respiratory rate the Group factor approached significance, F(1, 47)¼ 3.02, P¼0.089, Z2 ¼6.00% with high AS trending towards tonically elevated respiratory rate. For end-tidal pCO2 the Group Epoch interaction, F(3, 138)¼ 3.91, P¼0.048, Z2 ¼7.80% was significant. Inspections of the means in Fig. 3 suggest that this interaction was due to lower peak levels of pCO2 in high AS participants, indicative of less effective inhalations. However, post-hoc tests failed to locate this effect: between groups the t-test of peak epochs during CO2-B only approached significance, t(46)¼1.86, P¼0.070. Analysis of tidal volume showed a significant Group effect, F(1, 47)¼6.14, P¼0.017, Z2 ¼ 11.6%, due to lower values in the high AS group, in addition to a trend level

Group  Epoch interaction, F(3, 141) ¼2.77, P¼0.096, Z2 ¼7.53%. Inspection of the means in Fig. 3c suggests that this could be due to reduced peak values in high AS, however, exploratory post-hoc test failed to locate the trend-level interaction. Analysis of minute volume showed a trend level Group effect, F(1, 47)¼3.08, P¼0.086, Z2 ¼6.23%, due to slightly lower values in the high AS group. There were no other effects involving group in the other physiological measures.

4. Discussion The present study examined differential responding to CO2 provocations in healthy individuals with high and low AS, both on a macro-level (sensitization, habituation) and a micro-level (baseline, reactivity, recovery). We employed high temporal resolution dyspnea and anxiety assessment as well as broad respiratory and autonomic measurement. Our aim was to elucidate potential etiological mechanisms linking high AS in mentally healthy individuals to PD specifically and to clinical anxiety more generally. The results can be summarized as follows. Compared to low AS, high AS individuals showed tonically elevated anxiety ratings throughout all phases of the task. Reactivity was selectively increased on dyspnea ratings in high AS individuals during the post-CO2 air inhalations. HR showed a sensitization pattern, progressively increasing in high AS individuals, most notably during the second block of air inhalations (AIR-B). In the respiratory system, high AS individuals had tonically elevated respiratory rate, reduced tidal volume, and reduced peak pCO2 values during the CO2 inhalations. On a methodological level, these results demonstrate that the temporal dynamics of CO2 hyperreactivity in AS is characterized by changes at both the micro-level (within inhalations) and the macro-level (across inhalations), confirming the usefulness of our task design. Further, pre- and post-CO2 inhalations of room air

J. Blechert et al. / Psychiatry Research 209 (2013) 566–573

Minute volume (L/min)

18 16

high AS low AS

14 12 10 8 6

pCO2 (torr)

4 56 54 52 50 48 46 44 42 40 38 36

Tidal volume (liters)

1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2

PRE AIR-A

CO2-A

CO2-B

AIR-B

POST

Fig. 3. Means and standard errors of minute volume (a), end-tidal pCO2 (b), and tidal volume (c) for participants with high vs. low anxiety sensitivity. See Fig. 2 for details.

(AIR-A, AIR-B) revealed important aspects of AS reactivity as discussed below. In our previous study, specifically the PD group showed a nonhabituating, non-declining tidal volume course over the eight CO2 inhalations. The high AS group in the current study did not manifest this exact pattern, but two pieces of evidence confirmed a greater sensitivity of the respiratory system in individuals high in AS. First, respiratory rate was tonically elevated, and tidal volume attenuated in high AS compared to low AS individuals. The fact that these group differences were not related to specific phases of the task (i.e., the CO2 inhalations vs. the air inhalations) but were present throughout the task could be attributed to contextual anxiety in the laboratory context or to some stable difference between groups. Second, the combination of lower tidal volume across the whole task and lower peak CO2 levels suggests that there was some kind of behavioral avoidance of CO2 stimulation in high AS individuals. Thus, despite the explicit instructions that usually ensure reliable vital capacity inhalation, these subjects might not have inhaled as much CO2 as their low AS counterparts and might have generally reduced the depth of their breathing throughout the inhalations. Behavioral avoidance of CO2 has been noted in individuals at risk for PD (Coryell et al., 2001). Furthermore, Fannes et al. (2008) described a similar avoidance pattern in a respiratory conditioning task in healthy

571

individuals. Together these results suggest that like PD patients, high AS individuals show respiratory abnormalities although with a somewhat different profile: while PD patients fail to habituate to repeated CO2 stimulation, high AS individuals show a pattern of fast and shallow breathing, and seem to avoid CO2 stimulation. Despite these intriguing respiratory findings, clearly, similarities and differences in PD and high AS need more detailed study before drawing firm conclusions about a specific link between the two. In our previous study, both anxious groups showed tonically elevated anxiety and dyspnea responses throughout the task. Compared to low AS, high AS individuals evinced tonically elevated anxiety but there were no main effects for dyspnea. Also, high AS individuals did not show increased reactivity or delayed recovery during the CO2 inhalations, as had been observed for both clinical groups previously. Perhaps this reactivity/recovery pattern builds up when high AS individuals are repeatedly confronted with panic attacks or panic-like experiences. Or nonclinical anxiety may increasingly involve dyspnea and respiratory symptoms as it progresses to clinical anxiety. Similar to our previous study, the second set of air inhalation (AIR-B, after CO2 inhalations) discriminated AS groups: only high AS individuals showed increased dyspnea reactivity during peak air inhalation and this was accompanied by gradually increasing HR. Our previous study found this pattern for dyspnea in SP patients and increased anxiety ratings in PD. The fact that postCO2 exposure dyspnea reactivity was not as prominent before CO2 inhalation indicates that it is different from anticipatory or contextual anxiety. Dyspnea reactivity was confined to the peak of the air inhalation, making sensitization or incomplete recovery interpretations of this finding unlikely. Both of these mechanisms would have resulted in more general dyspnea elevations. Respiratory conditioning might have played a role here in that the inhalation procedure might have become associated with the effects of CO2 through the course of the inhalations, eventually leading to dyspneic sensations being triggered by the inhalation procedure, despite the fact that the inhaled gas mixture consisted of room air. This explanation, although tentative at this point since we did not use a differential conditioning design (cf., Acheson et al., 2007, 2012) fits well with the buildup of autonomic arousal over the task as indicated by increasing HR in the high AS group. Bouton et al. (2001) state that any vulnerability factor predisposing an individual to spontaneous panic attacks could, mediated by associative processes, increase the conditioning of external and internal cues to these attacks and result in more intense conditioned anxiety. As indicated above, this might be one etiological mechanism transforming high AS into clinical anxiety. Whereas enhanced conditionability has been demonstrated in PD and other anxiety disorders (Lissek et al., 2005; Blechert et al., 2007b; Michael et al., 2007), it is still a matter of controversy in AS (Forsyth et al., 1999). Our study was limited in several ways. First, our comparisons between high vs. low AS individuals were made cross-sectionally. Unequivocal evidence about etiological mechanisms would require replication in a longitudinal design. Second, even though most task parameters were identical, we did not study AS and PD/SP in the same study, which limits the comparability of results. Third, the fact that high AS individuals inhaled less CO2 might have led to an underestimation of group differences which – together with a small separation of high and low AS groups on the ASI (see next point) – could be the reason for the rather subtle respiratory group differences with small and partially trend level effects. Future studies might benefit from equalizing arterial CO2 dosages (which induce dyspnea and symptoms mainly through chemosensor feedback loops) across individuals. This might be accomplished by adjustment of CO2 dosages or reinforcement of inhalation

572

J. Blechert et al. / Psychiatry Research 209 (2013) 566–573

instructions, both guided by on-line monitoring of pCO2 values. Fourth, future studies might profit from comparing AS groups that were formed according to the taxonic properties of the ASI (Bernstein et al., 2005, 2007, but see also Asmundson et al., 2011; Bernstein et al., 2010; Broman-Fulks et al., 2010) and to control for differences in trait anxiety using an orthogonalizing recruitment procedure (essentially including a sufficient number of participants with high scores on one vs. low scores on the other questionnaire) to disentangle the relative contributions of the two constructs (e.g., Rapee and Medoro, 1994; Carter et al., 2001). Recruitment of a demographically more diverse sample would increase the generalizability of such findings. In summary, the present task has revealed psychological and psychophysiological response patterns that help to understand the role of high AS as a precursor for clinical anxiety and for PD in particular. The findings provide evidence to support preventive action targeted at high AS individuals (Schmidt et al., 2007). Our task also might be usefully applied as an interoceptive exposure therapy for panic disorder (cf., Beck et al., 1997). Modification of both within-inhalation and across-inhalation processes could be important therapeutically according to a recent discussion about within- and between-session habituation for successful exposure therapy in anxiety disorders (Craske et al., 2008).

Acknowledgments This research was supported by the Department of Veterans Affairs (WTR), NIH Grant MH56094 (WTR, FHW), and the Swiss National Science Foundation Grant 105311-105850 (FHW). We thank Johannes Thiermann for his help with the data collection.

References Abrams, K., Rassovsky, Y., Kushner, M.G., 2006. Evidence for respiratory and nonrespiratory subtypes in panic disorder. Depression and Anxiety 23, 474–481. Acheson, D.T., Forsyth, J.P., Moses, E., 2012. Interoceptive fear conditioning and panic disorder: the role of conditioned stimulus-unconditioned stimulus predictability. Behavior Therapy 43, 174–189. 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% CO(2)-enriched air in a non-clinical sample. Behaviour Research and Therapy 45, 2280–2294. Ainsworth, B.E., Bassett Jr., D.R., Strath, S.J., Swartz, A.M., O’Brien, W.L., Thompson, R.W., Jones, D.A., Macera, C.A., Kimsey, C.D., 2000. Comparison of three methods for measuring the time spent in physical activity. Medicine and Science in Sports and Exercise 32, S457–464. American Psychiatric Association, A.P.A., 1994. Diagnostic and Statistical Manual of Mental Disorders, fourth edition American Psychiatric Association, Washington, DC. Asmundson, G.J., Norton, G.R., Wilson, K.G., Sandler, L.S., 1994. Subjective symptoms and cardiac reactivity to brief hyperventilation in individuals with high anxiety sensitivity. Behaviour Research and Therapy 32, 237–241. Asmundson, G.J., Taylor, S., 1996. Role of anxiety sensitivity in pain-related fear and avoidance. Journal of Behavioral Medicine 19, 577–586. Asmundson, G.J.G., Weeks, J.W., Nicholas Carleton, R., Thibodeau, M.A., Fetzner, M.G., 2011. Revisiting the latent structure of the anxiety sensitivity construct: more evidence of dimensionality. Journal of Anxiety Disorders 25, 138–147. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J.E., Erbaugh, J.K., 1961. An inventory for measuring depression. Archives of General Psychiatry 4, 561–571. Beck, J.G., Shipherd, J.C., 1997. Repeated exposure to interoceptive cues: does habituation of fear occur in panic disorder patients? A preliminary report. Behavior Research and Therapy 35, 551–557. Beck, J.G., Shipherd, J.C., Read, J., 1999. Response patterns to repeated CO2 inhalation in individuals with high anxiety sensitivity. Behaviour Research and Therapy 37, 1073–1089. Beck, J.G., Shipherd, J.C., Zebb, B.J., 1996. Fearful responding to repeated CO2 inhalation: a preliminary investigation. Behaviour Research and Therapy 34, 609–620. Beck, J.G., Shipherd, J.C., Zebb, B.J., 1997. How does interoceptive exposure for panic disorder work? An uncontrolled case study. Journal of Anxiety Disorders 11, 541–556.

Beck, J.G., Wolf, M.S., 2001. Response to repeated CO2 in individuals with elevated anxiety sensitivity: replication with 20% CO2. Journal of Behavior Therapy and Experimental Psychiatry 32, 1–16. Bernstein, A., Stickle, T.R., Zvolensky, M.J., Taylor, S., Abramowitz, J., Stewart, S., 2010. Dimensional, categorical, or dimensional-categories: testing the latent structure of anxiety sensitivity among adults using factor-mixture modeling. Behavior Therapy 41, 515–529. Bernstein, A., Zvolensky, M.J., Feldner, M.T., Lewis, S.F., Fauber, A.L., Leen-Feldner, E.W., Vujanovic, A.A., 2005. Anxiety sensitivity taxon and trauma: discriminant associations for posttraumatic stress and panic symptomatology among young adults. Depression and Anxiety 22, 138–149. Bernstein, A., Zvolensky, M.J., Norton, P.J., Schmidt, N.B., Taylor, S., Forsyth, J.P., Lewis, S.F., Feldner, M.T., Leen-Feldner, E.W., Stewart, S.H., Cox, B., 2007. Taxometric and factor analytic models of anxiety sensitivity: integrating approaches to latent structural research. Psychological Assessment 19, 74–87. Blechert, J., Michael, T., Grossman, P., Lajtman, M., Wilhelm, F.H., 2007a. Autonomic and respiratory characteristics of posttraumatic stress disorder and panic disorder. Psychosomatic Medicine 69, 935–943. Blechert, J., Michael, T., Vriends, N., Margraf, J., Wilhelm, F.H., 2007b. Fear conditioning in posttraumatic stress disorder: evidence for delayed extinction of autonomic, experiential, and behavioural responses. Behaviour Research and Therapy 45, 2019–2033. Blechert, J., Wilhelm, F.H., Meuret, A.E., Wilhelm, E.M., Roth, W.T., 2010. Respiratory, autonomic, and experiential responses to repeated inhalations of 20% CO enriched air in panic disorder, social phobia, and healthy controls. Biological Psychology 84, 104–111. Bouton, M.E., Mineka, S., Barlow, D.H., 2001. A modern learning theory perspective on the etiology of panic disorder. Psychological Review 108, 4–32. Broman-Fulks, J.J., Deacon, B.J., Olatunji, B.O., Bondy, C.L., Abramowitz, J.S., Tolin, D.F., 2010. Categorical or dimensional: a reanalysis of the anxiety sensitivity construct. Behavior Therapy 41, 154–171. Burkhardt, S.C.A., Wilhelm, F.H., Meuret, A.E., Blechert, J., Roth, W.T., 2010. Temporal stability and coherence of anxiety, dyspnea, and physiological variables in panic disorder. Biological Psychology 85, 226–232. Caldirola, D., Perna, G., Arancio, C., Bertani, A., Bellodi, L., 1997. The 35% CO2 challenge test in patients with social phobia. Psychiatry Research 71, 41–48. Carter, M.M., Suchday, S., Gore, K.L., 2001. The utility of the ASI factors in predicting response to voluntary hyperventilation among nonclinical participants. Journal of Anxiety Disorders 15, 217–230. Coryell, W., Fyer, A., Pine, D., Martinez, J., Arndt, S., 2001. Aberrant respiratory sensitivity to CO(2) as a trait of familial panic disorder. Biological Psychiatry 49, 582–587. Craske, M.G., Kircanski, K., Zelikowsky, M., Mystkowski, J., Chowdhury, N., Baker, A., 2008. Optimizing inhibitory learning during exposure therapy. Behaviour Research and Therapy 46, 5–27. Dager, S.R., Strauss, W.L., Marro, K.I., Richards, T.L., Metzger, G.D., Artru, A.A., 1995. Proton magnetic resonance spectroscopy investigation of hyperventilation in subjects with panic disorder and comparison subjects. American Journal of Psychiatry 152, 666–672. Deacon, B., Abramowitz, J., 2006. Anxiety sensitivity and its dimensions across the anxiety disorders. Journal of Anxiety Disorders 20, 837–857. Fannes, S., Van Diest, I., Meulders, A., De Peuter, S., Vansteenwegen, D., Van den Bergh, O., 2008. To inhale or not to inhale: conditioned avoidance in breathing behavior in an odor—20% CO2 paradigm. Biological Psychology 78, 87–92. Finlay, C.G., Forsyth, J.P., 2009. Context and renewal of conditioned fear: an experimental evaluation using 20% carbon dioxide-enriched air as an unconditioned stimulus. Journal of Anxiety Disorders 23, 737–745. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1995. Structured Clinical Interview for DSM-IV Axis I Disorders. American Psychiatric Press, Washington, D.C. 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, 291–304. Forsyth, J.P., Palav, A., Duff, K., 1999. The absence of relation between anxiety sensitivity and fear conditioning using 20% versus 13% CO-sub-2-enriched air as unconditioned stimuli. Behaviour Research and Therapy 37, 143–153. Freire, R.C., Lopes, F.L., Valenc- a, A.M., Nascimento, I., Veras, A.B., Mezzasalma, M.A., de-Melo-Neto, V.L., Zin, W.A., Nardi, A.E., 2008. Panic disorder respiratory subtype: a comparison between responses to hyperventilation and CO2 challenge tests. Psychiatry Research 157, 307. Freire, R.C., Perna, G., Nardi, A.E., 2010. Panic disorder respiratory subtype: psychopathology, laboratory challenge tests, and response to treatment. Harvard Review of Psychiatry 18, 220–229. Gorman, J.M., Martinez, J., Coplan, J.D., Kent, J., Kleber, M., 2004. The effect of successful treatment on the emotional and physiological response to carbon dioxide inhalation in patients with panic disorder. Biological Psychiatry 56, 862–867. Gorman, J.M., Papp, L.A., Martinez, J., Goetz, R.R., Hollander, E., Liebowitz, M.R., Jordan, F., 1990. High-dose carbon dioxide challenge test in anxiety disorder patients. Biological Psychiatry 28, 743–757. Grassi, M., Caldirola, D., Vanni, G., Guerriero, G., Piccinni, M., Valchera, A., Perna, G., 2013. Baseline respiratory parameters in panic disorder: a meta-analysis. Journal of Affective Disorders 146, 158–173. Griez, E., Deloof, C., Pols, H., Zandbergen, J., Lousberg, H., 1990. Specific sensitivity of patients with panic attacks to carbon dioxide inhalation. Psychiatry Research 31, 193–199.

J. Blechert et al. / Psychiatry Research 209 (2013) 566–573

Hayward, C., Killen, J.D., Kraemer, H.C., Taylor, C.B., 2000. Predictors of panic attacks in adolescents. Journal of the American Academy of Child and Adolescent Psychiatry 39, 207–214. Holloway, W., McNally, R.J., 1987. Effects of anxiety sensitivity on the response to hyperventilation. Journal of Abnormal Psychology 96, 330–334. Joiner, T.E., Anestis, M.D., Holm-Denoma, J.M., Gordon, K.H., Schmidt, N.B., 2008. The role of anxiety sensitivity in eating pathology. Cognitive Therapy and Research 32, 370–385. Kelly, M.M., Forsyth, J.P., 2009. Associations between emotional avoidance, anxiety sensitivity, and reactions to an observational fear challenge procedure. Behaviour Research and Therapy 47, 331–338. Kilic, E.Z., Kilic, C., Yilmaz, S., 2008. Is anxiety sensitivity a predictor of PTSD in children and adolescents? Journal of Psychosomatic Research 65, 81–86. Killgore, W.D.S., Britton, J.C., Price, L.M., Gold, A.L., Deckersbach, T., Rauch, S.L., 2011. Neural correlates of anxiety sensitivity during masked presentation of affective faces. Depression and Anxiety 28, 243–249. Kircanski, K., Craske, M.G., Epstein, A.M., Wittchen, H.U., 2011. Subtypes of panic attacks: a critical review of the empirical literature. Focus 9, 389. Klein, D.F., 1993. False suffocation alarms, spontaneous panics, and related conditions. An integrative hypothesis. Archives of General Psychiatry 50, 306–317. Klein, D.F., 1994. Testing the suffocation false alarm theory of panic disorder. Anxiety 1, 1–7. Lang, A.J., Kennedy, C.M., Stein, M.B., 2002. Anxiety sensitivity and PTSD among female victims of intimate partner violence. Depression and Anxiety 16, 77–83. Ley, R., 1985. Blood, breath, and fears: a hyperventilation theory of panic attacks and agoraphobia. Clinical Psychology Review 5, 271–285. Li, W., Daems, E., Van de Woestijne, K.P., Van Diest, I., Gallego, J., De Peuter, S., Bogaerts, K., Van den Bergh, O., 2006. Air hunger and ventilation in response to hypercapnia: effects of repetition and anxiety. Physiology and Behavior 88, 47–54. Li, W., Van Diest, I., De Peuter, S., Bogaerts, K., Oyen, N., Hombroux, N., Van de Woestijne, K., Gallego, J., Van den Bergh, O., 2008. Repeated experiences of air hunger and ventilatory behavior in response to hypercapnia in the standardized rebreathing test: effects of anxiety. Biological Psychology 77, 223–232. Lissek, S., Powers, A.S., McClure, E.B., Phelps, E.A., Woldehawariat, G., Grillon, C., Pine, D.S., 2005. Classical fear conditioning in the anxiety disorders: a metaanalysis. Behaviour Research and Therapy 43, 1391–1424. Maddock, R.J., Carter, C.S., 1991. Hyperventilation-induced panic attacks in panic disorder with agoraphobia. Biological Psychiatry 29, 843–854. Maller, R.G., Reiss, S., 1992. Anxiety sensitivity in 1984 and panic attacks in 1987. Journal of Anxiety Disorders 6, 241–247. Marshall, G.N., Miles, J.N., Stewart, S.H., 2010. Anxiety sensitivity and PTSD symptom severity are reciprocally related: evidence from a longitudinal study of physical trauma survivors. Journal of Abnormal Psychology 119, 143–150. McMillan, K.A., Asmundson, G.J.G., Zvolensky, M.J., Carleton, R.N., 2012. Startle response and anxiety sensitivity: subcortical indices of physiologic arousal and fear responding. Emotion 12, 1264–1272. McWilliams, L.A., Becker, E.S., Margraf, J., Clara, I.P., Vriends, N., 2007. Anxiety disorder specificity of anxiety sensitivity in a community sample of young women. Personality and Individual Differences 42, 345–354. Melzig, C.A., Holtz, K., Michalowski, J.M., Hamm, A.O., 2011. Interoceptive threat leads to defensive mobilization in highly anxiety sensitive persons. Psychophysiology 48, 745–754. Meuret, A.E., Ritz, T., 2010. Hyperventilation in panic disorder and asthma: empirical evidence and clinical strategies. International Journal of Psychophysiology 78, 68–79. Meuret, A.E., Rosenfield, D., Wilhelm, F.H., Zhou, E., Conrad, A., Ritz, T., Roth, W.T., 2011. Do unexpected panic attacks occur spontaneously? Biological Psychiatry 70, 985–991. Michael, T., Blechert, J., Vriends, N., Margraf, J., Wilhelm, F.H., 2007. Fear conditioning in panic disorder: enhanced resistance to extinction. Journal of Abnormal Psychology 116, 612–617. Miller, G.A., Chapman, J.P., 2001. Misunderstanding analysis of covariance. Journal of Abnormal Psychology 110, 40–48. Naragon-Gainey, K., 2010. Meta-analysis of the relations of anxiety sensitivity to the depressive and anxiety disorders. Psychological Bulletin 136, 128–150. Niccolai, V., Van Duinen, M., Griez, E., 2009. Respiratory patterns in panic disorder reviewed: a focus on biological challenge tests. Acta Psychiatrica Scandinavica 120, 167–177. Olatunji, B.O., Wolitzky-Taylor, K.B., 2009. Anxiety sensitivity and the anxiety disorders: a meta-analytic review and synthesis. Psychological Bulletin 135, 974–999.

573

Olatunji, B.O., Wolitzky-Taylor, K.B., Elwood, L., Connolly, K., Gonzales, B., Armstrong, T., 2009. Anxiety sensitivity and health anxiety in a nonclinical sample: specificity and prospective relations with clinical stress. Cognitive Therapy and Research 33, 416–424. Papp, L.A., Martinez, J.M., Klein, D.F., Coplan, J.D., Norman, R.G., Cole, R., de Jesus, M.J., Ross, D., Goetz, R., Gorman, J.M., 1997. Respiratory psychophysiology of panic disorder: three respiratory challenges in 98 subjects. American Journal of Psychiatry 154, 1557–1565. Pelosi, P., Croci, M., Ravagnan, I., Tredici, S., Pedoto, A., Lissoni, A., Gattinoni, L., 1998. The effects of body mass on lung volumes, respiratory mechanics, and gas exchange during general anesthesia. Anesthesia and Analgesia 87, 654–660. Perna, G., Romano, P., Caldirola, D., Cucchi, M., Bellodi, L., 2003. Anxiety sensitivity and 35% CO2 reactivity in patients with panic disorder. Journal of Psychosomatic Research 54, 573–577. Pfaltz, M.C., Michael, T., Grossman, P., Blechert, J., Wilhelm, F.H., 2009. Respiratory pathophysiology of panic disorder: an ambulatory monitoring study. Psychosomatic Medicine 71, 869–876. Rapee, R.M., Medoro, L., 1994. Fear of physical sensations and trait anxiety as mediators of the response to hyperventilation in nonclinical subjects. Journal of Abnormal Psychology 103, 693–699. Rassovsky, Y., Kushner, M.G., 2003. Carbon dioxide in the study of panic disorder: issues of definition, methodology, and outcome. Journal of Anxiety Disorders 17, 1–32. Reiss, S., 1991. Expectancy model of fear, anxiety, and panic. Clinical Psychology Review 11, 141–153. 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. Roberson-Nay, R., Latendresse, S., Kendler, K., 2012. A latent class approach to the external validation of respiratory and non-respiratory panic subtypes. Psychological Medicine 42, 461. Rodriguez, B.F., Bruce, S.E., Pagano, M.E., Spencer, M.A., Keller, M.B., 2004. Factor structure and stability of the Anxiety Sensitivity Index in a longitudinal study of anxiety disorder patients. Behaviour Research and Therapy 42, 79–91. Schmidt, N.B., Eggleston, A.M., Woolaway-Bickel, K., Fitzpatrick, K.K., Vasey, M.W., Richey, J.A., 2007. Anxiety Sensitivity Amelioration Training (ASAT): a longitudinal primary prevention program targeting cognitive vulnerability. Journal of Anxiety Disorders 21, 302–319. Schmidt, N.B., Zvolensky, M.J., Maner, J.K., 2006. Anxiety sensitivity: prospective prediction of panic attacks and Axis I pathology. Journal of Psychiatric Research 40, 691–699. Smits, J.A., Berry, A.C., Tart, C.D., Powers, M.B., 2008. The efficacy of cognitivebehavioral interventions for reducing anxiety sensitivity: a meta-analytic review. Behaviour Research and Therapy 46, 1047–1054. Spielberger, C.D., Gorsuch, R.L., Luchene, R.E., 1970. State-Trait Anxiety Inventory. Consulting Psychology Press, Palo Alto. Taylor, S., Zvolensky, M.J., Cox, B.J., Deacon, B., Heimberg, R.G., Ledley, D.R., Abramowitz, J.S., Holaway, R.M., Sandin, B., Stewart, S.H., Coles, M., Eng, W., Daly, E.S., Arrindell, W.A., Bouvard, M., Cardenas, S.J., 2007. Robust dimensions of anxiety sensitivity: development and initial validation of the Anxiety Sensitivity Index-3. Psychological Assessment 19, 176–188. Telch, M.J., Harrington, P.J., Smits, J.A., Powers, M.B., 2011. Unexpected arousal, anxiety sensitivity, and their interaction on CO(2)-induced panic: further evidence for the context-sensitivity vulnerability model. Journal of Anxiety Disorders 25, 645–653. van den Hout, M.A., van der Molen, G.M., Griez, E., Lousberg, H., Nansen, A., 1987. Reduction of CO2-induced anxiety in patients with panic attacks after repeated CO2 exposure. American Journal of Psychiatry 144, 788–791. Vickers, K., 2012. Hypersensitivity to hypercapnia: definition/(s). Psychiatry Research 197, 7–12. Wilhelm, F.H., Gerlach, A.L., Roth, W.T., 2001a. Slow recovery from voluntary hyperventilation in panic disorder. Psychosomatic Medicine 63, 638–649. Wilhelm, F.H., Grossman, P., Roth, W.T., 2005. Assessment of heart rate variability during alterations in stress: complex demodulation vs. spectral analysis. Biomedical Sciences Instrumentation 41, 346–351. Wilhelm, F.H., Peyk, P., 2005. Autonomic Nervous System Laboratory (ANSLAB)—Shareware Version. Software Repository of the Society for Psychophysiological Research. /http://www.sprweb.orgS. Wilhelm, F.H., Trabert, W., Roth, W.T., 2001b. Physiological instability in panic disorder and generalized anxiety disorder. Biological Psychiatry 49, 596–605.