Effects of induced hyperventilation on electrodermal response habituation to agoraphobia — Relevant stimuli

Journal of Psychosomatic Pnnted in Great Britain.

Research,

EFFECTS

Vol. 31, No. 3, pp. 401412.

OF INDUCED

ELECTRODERMAL AGORAPHOBIA

0022.3999/87 $3.00 + .@I Pergamon Journals Ltd.

1987.

HYPERVENTILATION

RESPONSE HABITUATION - RELEVANT STIMULI

ON TO

LUKE D. KARTSOUNIS* and GRAHAM TURPIN~ (Received 15 September

1986; accepted in revised form 23 January 1987)

Abstract-The role of hyperventilation in the aetiology of anxiety disorders was investigated in an analogue study. It was hypothesised that induced hyperventilation would alter subsequent subjective and physiological responses to visually presented agoraphobic material. Three groups (N = 16) of students were used and each was divided into two sub-groups which were presented with slides of either potentially agoraphobic or neutral content. During the first phase of the experiment, subjects were exposed to 10 slides whilst their breathing rate was manipulated. Group FB was instructed to hyperventilate by asking the subjects to breathe fast and deeply, Group SB was instructed to breathe at a slow rate and Group NB breathed normally. After a recovery phase, subjects were re-exposed to the same visual stimuli as used in the first phase of the experiment. Measures of skin conductance response amplitude were obtained for each stimulus in the series. In addition, heart rate and respiration were monitored, and subjective meassres of somatic and affective states were obtained. Subjects also rated the pleasantness of the stimuli. The hyperventilation instructions resulted in delayed electrodermal habituation relative to the other groups. This finding could not be accounted for in terms of group differences in either physiological or subjective ‘arousal’.

INTRODUCTION

HYPERVENTILATION (HV) is defined as an increase in pulmonary ventilation greater than that required for the metabolic needs of the body. It results in a reduction of the partial pressure of arterial carbon dioxide levels (pCOz) and is expressed as various subjective symptoms [l]. It often arises from rapid ventilation or over-breathing and has commonly been associated with anxiety. For example, symptoms of both anxiety and the so-called Hyperventilation Syndrome (HVS) are similar (i.e. dizziness, sweating, palpitations, apprehension, panic), and the traditional view has been that HV is caused by anxiety. This claim has been supported by clinical [2, 31 and experimental studies [4] which have shown that anxiety may indeed result directly to overbreathing and HV. The opposing view has also been advanced, whereby anxiety is considered to arise from HV. Lum [5] has argued that HV is ‘essentially a bad habit’ and that it can be primary to anxiety symptoms. Nevertheless, Lum [6] conceded that a variety of psychological and somatic antecedents can initiate a positive feedback loop whereby HV results in anxiety symptoms which lead to continued HV and the further exacerbation of symptoms. A similar model was advanced by Lewis [7]. This implies that anxiety may be both the cause and consequence of HV. Some support for Lum’s ideas may be derived from clinical interventions which purport to alleviate certain anxiety symptoms. For example, anxiety is said to be relieved both by breathing COz-enriched air [S, 91 and rebreathing expired air from *Department tDepartment

of Psychology. of Psychology,

The National Hospital Plymouth Polytechnic. 401

and the Brook

General

Hospital,

London

402

L. D. KARTSOUNIS

and G. TURPIN

a bag [lo]. Both these techniques result in an increase in pCO? which is said to underlie HV symptoms. Such practices, however, do not provide direct evidence of the causation of anxiety symptoms since methods of symptom-relief may not be directly related to aetiology. Furthermore, Van den Hout and Griez [ll] argued that CO2 inhalation itself produces no direct anxiolytic effect but the overall sub,jective experience is dependent upon the subjects’ expectations. Several other writers have also discussed the potential role of HV in the aetiology of anxiety disorders, and particularly that of agoraphobia and panic disorder. It is generally thought that one of the major characteristics of agoraphobia is fear of panic attacks. Indeed, panic attacks are commonly reported by agoraphobic clients as the origin of their difficulties, and several clinicians [12-141 have suggested that panic may have initially occurred as a result of HV. Recently, Garssen, Van Veenendaal and Bloemink [15] found a substantial overlap between HV and agoraphobic syndromes (over 60%) which they interpreted as supporting an aetiological role of HV in the development of agoraphobia. Since this was essentially a correlational study, the direction of the association between the two syndromes remains uncertain. Some further support for the role of HV in panic disorder has been demonstrated in rescent treatment studies which have employed respiratory training [16, 171. Similarly, Ley [I81 advocated voluntary breathing control for the treatment of agoraphobia and other disorders related to fear of panic. The present research was conducted in an attempt to examine experimentally the nature of the relationship between agoraphobia and HV. An analogue study was designed based upon the electrodermal response habituation paradigm which has been employed as a basis for studying other anxiety-related problems [lo, 201. It was postulated that if the aversive effects of HV have a primary role in the aetiology of anxiety problems it would be expected that relatively innocuous stimuli presented contiguously with induced HV will become negatively evaluated [21], and they will acquire increased attentional value which will be reflected in the magnitude and habituation rate of the orienting response [22]. It was specifically predicted, therefore, that electrodermal responses elicited by stimuli associated with induced HV will be larger and display slower habituation than responses to the same stimuli initially presented in the context of either normal or slow breathing. It was further hypothesised that the difference in habituation would be more prominent for agoraphobia-relevant than for neutral material. Finally, it has been suggested by Mathews, Gelder and Johnson [23] that persons who score as ‘external’ on Rotter’s [24] Locus of Control scale may be more prone to developing agoraphobia than internal locus of control individuals. Accordingly, the effects of Locus of Control on electrodermal habituation to agoraphobia-relevant stimuli were explored. METHOD

Three groups of subjects were employed. and each group was divided into two sub-groups which were presented with slides of either potentially agoraphobic or supposedly neutral content. In the first phase of the experiment, subjects viewed their respective slides and were asked to comply with a \et of tape-recorded verbal instructions. Group FB (Fast Breathing) were required to breath fast and deeply, group SB (Slow Breathing) were asked to breath slowly and group NB (Normal Breathing) received no instructions which directly referred to their breathing rate. After a recovery period, the

Hyperventilation

and electrodermal

403

habituation

second phase of the experiment commenced when subjects viewed again their respective stimuli presented in the form of an electrodermal habituation paradigm. The major dependent variables were the magnitude of the skin conductance responses elicited by the slides presented in the second phase of the study and subjective ratings of the pleasantness-unpleasantness of the stimuli. In addition, the effects of the breathing manipulation upon induced HV were indirectly assessed using both tonic measures of skin conductance and heart rate, together with subjective questionnaires. Subjects A total of 66 first-year psychology undergraduates participated in partial fulfilment of a course requirement. Fifteen of these subjects were excluded because of apparatus failure or experimenter’s error. A further three subjects were dropped from the FB group due to their failure to comply with the instructions to hyperventilate. Hence, 48 subjects were included in the study, 16 in each group. The mean age was 20.5 yr (range l&37), and there were 36 females and 12 males. Selection of stimulus material In selecting potentially agoraphobic stimuli, two criteria were applied: first, the stimuli should depict situations which agoraphobic clients tend to fear and avoid; second, they should be easily represented in pictorial form. For the neutral stimuli the criteria were that they should be similar to the agoraphobic ones in general characteristics such as colour, complexity and brightness but should not be either fear relevant or especially attractive. A series of slides, as opposed to a single slide. was employed to ensure that the ‘agoraphobic’ content was generally portrayed. Thirty coloured photographs of agoraphobia-relevant scenes were taken, and a sample of the ten most representative were chosen by three clinical psychologists. These were then rated by three agoraphobic clients in terms of the anxiety they evoked on a 1 (‘not at all’) to 5 (‘very much’) scale. Three of the scenes were rated by all three patients as feared ‘a fair amount’ or above (mean rating = 3.66). These scenes had the same general content which portrayed streets with crowds and cars. Two more scenes with the same content were prepared, and the five photographs were made into slides. Five neutral stimuli were also selected from 30 photographs of flowers and garden scenes by the same clinical psychologists. In order to validate further the stimulus material, all ten slides were rated by another three agoraphobic clients. All the agoraphobic stimuli were rated as either ‘much feared’ or ‘very much feared’ (mean rating = 4.67), whereas the neutral stimuli were rated as either ‘not at all’ or ‘a little’ (mean rating = 1.33). The final stimulus series consisted of ten slides composed of either five neutral or five agoraphobic slides. each presented twice. Apparatus Subjects were seated in a comfortable chair in a dimly lit, temperature controlled and sound-attenuated room..The stimulus equipment and recording apparatus were-operated from an adjoining room. The slides (34 X 23 mm) were nroiected from a oroiector (Kodak Carousel. S-AV 2020) which was controlled by a slide synchronizer (Eh;l 1920) and audio~visualstero-cassette recorder (Phillips, D 6920). They were back-projected on a glass screen at approximately 1.5 m from the subject’s forehead. The size of the visible picture was 150 X 210 mm. When the stimulus slides were not projected, the background was illuminated by a continuous blue slide. The slide synchronizer was also employed to present via loud-speakers the verbal instructions to the subjects in the first phase of the experiment. Skin conductance, respiration and heart rate were monitored using a Grass Model 7 polygraph. Skin conductance was recorded using a constant-voltage technique [25], and bipolar Ag/AgCl electrodes (9 mm diameter) attached by self-adhesive electrode collars to the distal phalanges of the first and second fingers of the subject’s non-preferred hand. Respiration was derived from a strain gauge secured around the subject’s chest. The electrocardiogram was measured using a Grass 7 P6 preamplifier, and plate electrodes attached to the subject’s wrists. The output from the driver amplifier was used to trigger a Grass 7 P4D cardiotachometer in order to obtain a beat-by-beat record of heart rate. The same electrolyte jelly (Johnson & Johnson, KY) was employed in order to measure both skin conductance and heart rate. Subjects were required to complete a series of questionnaires which included: the Symptom Rating Sheet (SRS) - a self-report scale of HV [26]; the General Arousal Scale (GAS) - a scale of subjective arousal derived from Thayer’s Activation-Deactivation Adjective Checklist [27]; the Locus of Control Questionnaire [24] and a seven-point rating scale of the pleasantness-unpleasantness of the stimulus material. Procedure Upon arrival

at the laboratory,

subjects

were asked

whether

they suffered

from respiratory,

heart,

404

L. D. KARTSXJNIS

and G. TURPIN

or epileptic problems. None of the participants reported any such disorder and they all confirmed that they had not smoked or exercised strenuously during the 15 min prior to their arrival. After being seated in the experimental room, they were told that the purpose of the study was to investigate physiological reactivity to some simple tasks. However, they were all warned that during the experiment they might feel uncomfortable and could withdraw from the experiment. A consent form was signed by all subjects. The physiological transducers were attached to the subjects and their activity was monitored. After 3 min. the subjects completed the SRS and GAS questionnaires and were given the first set of experimental instructions. They were told that they would be presented with some slides and that they should look at the contents of each slide since they would be asked some questions about them later. It was explained, however, that this was not a memory test. Subjects in groups FB and SB were asked to breath in deeply via the mouth and nose every time they heard ‘in’ and to breathe out every time they heard ‘out’. Subjects in group NB were not instructed to alter their breathing. However, in order to control for the attentional demands imposed on groups FB and SB, the subjects in group NB were asked to ensure that they listened to a comment from the speakers repeating ‘one-two’ throughout the presentation of the slides. Subjects were presented with 10 slides. each for 6.5 sec. During the presentation, group FB was instructed to breath at a rate of 30 cycles per min. Compliance with these instructions was defined in terms of the required breathing rate, as well as a substantial increase in the amplitude of respiration for at least a third of the 2 min session. Group SB was instructed to breath at a rate of 6 cycles per min. Their compliance was defined in terms of the required breathing rate. Subjects in group NB listened to the comment ‘one-two’ at a rate of 20 cycles per min. At the end of the first phase of the experiment, subjects were again requested to complete the SRS and GAS questionnaires. Following this. they were given a 10 min recovery period and the second set of instructions. They were told that they would be presented once more with slides but they would no longer be required to listen to or comply with any comment given over the loud-speakers. Each sub-group received their respective stimulus series as in the first phase of the study but presented in random order using an habituation paradigm. Each slide was shown for 6.5 set and the inter-stimulus interval ranged from 20 to 34 set with a mean interval of 25 sec. Upon the completion of the habituation series, subjects again completed the SRS and GAS questionnaires. They were also asked to rate both the neutral and agoraphobic stimuli in terms of pleasantness-unpleasantness. In this instance, the stimuli were presented randomly in photographic form, each for 2 sec. Finally the Locus of Control Questionnaire was administered and the subjects were debriefed. The criterion of a skin conductance response (SCR) was dcfincd as an increase in skin conductance of at least 0.01 pmhos. This was estimated as the maximal change in conductance between prestimulus and stimulus offset levels. Responses with latencies of 2.5 set or greater were not counted. Habituation of SCRs was measured by three methods: decrease in response magnitude over trial\, response probability and the number of stimulus presentations before three consecutive zero (i.e. below criterion) responses. Phasic SCRs here only measured in the second (habituation) phase of the experiment since responses in the first phase of the study were considered unscorable due to the breathing manipulation and the short inter-stimulus intervals employed. Four measures of tonic physiological aclivity were obtained and these included: skin conductance level (SCL). non-specific skin conductance responses (NSCRs), heart rate (HR) and respiratory rate (RR). Skin conductance level was assessed at minute intervals. NSCRs were counted for each minute, HR was derived from the mean of five consecutive inter-beat intervals which occurred at each minute. and RR was derived from the time measured between two successive inspiratory peaks at each minute. These measures were obtained throughout the experiment and were yampled before the first phase. during the breathing manipulation, after the recovery period and during the habituation series. RESULTS

Physiological and self-report data were subjected to analysis of variance (ANOVA) with repeated measures. Conservative degrees of freedom [28] were employed and a 0.05 confidence level was applied to all statistical tests. Analysis

of pre-manipulation

measures

In order to test for pre-experimental differences between the groups and subgroups, two-way ANOVAs with Breathing Manipulation (3 levels) and Slide Con-

Hyperventilation

tent (2 levels) were carried and the SRS. No significant

and electrodermal

habituation

405

out on the tonic physiological measures and the GAS main effects were observed for these factors.

Analysis of breathing manipulation phase measures Two aspects of the breathing manipulation phase of the study were investigated. First, differences in respiratory rate across groups FB, SB and NB were assessed in order to substantiate the effects of the breathing instructions. This was accomplished using a three-way ANOVA with Breathing Manipulation (3 levels), Slide Content (2 levels) and Minutes (2 levels) as the factors. A significant main effect only for Breathing Manipulation (F(2, 42) = 441.3, MSe = 1.279) was obtained. The mean RR for the FB, SB and NB groups were 31.09, 6.02 and 15.71 cycles per min respectively. A planned comparison between FB and the other two groups was significant (F(1, 42) = 205.4). These data are displayed in Fig. 1. Second, comparisons were made across the groups using both tonic autonomic and subjective measures in order to provide indirect evidence that overbreathing in the FB group gave rise to HV. More specifically, it was predicted that HV as a result of the instructions given to the FB group would result in elevated HR and SCL, and increased self-report of HV symptoms on the SRS. Similar three-way ANOVAs as above were employed. For HR, only a significant effect for Breathing Manipulation (F(2, 42) = 5.92, MSe = 721.5) was revealed. A planned comparison between FB and the two other groups was significant (F(1, 42) = 5.91), indicating that HR was elevated for this group. For SCL, neither the groups or sub-groups factors were significant. However, a significant Minutes factor (F(1, 42) = 17.9), decline in maniMSe = 0.0006) indicated a SCL during the pulation phase. Somatic and affective scores were extracted from the SRS [26] and analysed separately. Significant effects for Breathing Manipulation were obtained for somatic and affective scores (F(1, 42) = 10.05 and 4.28, MSe = 0.11 and 0.28) respectively. Planned comparisons indicated that the FB group differed significantly from the other two groups on both the somatic and affective scores (F(1, 42) = 12.73 and 5.50) respectively. These data are shown in Figs 1 and 2. In order to explore the possibility that the breathing instructions might have led to differential effects on cognitive arousal the GAS was employed. However, ANOVA revealed no significant effects (F(1, 42)
L. D. KARTSOUNIS

406

and G. TURPIN

Fast breathing group . Slow breathmg group A Normal breathing group l

(a)

10

120

r

(b)

r

Cd)

FIG. 1. -Mean change in (a) skin conductance level (SCL); (b) frequency of non-specific skin conductance responses (NSCRs); (c) heart rate (HR) and (d) respiratory rate (RR) throughout experimental periods: (1) pre-breathing manipulation; (2) breathing manipulation; (3) recovery and (4) habituation.

Hyperventilation n Fast

breathing group

(b)

SRS (Somatic)

(cl

and electrodermal 0 Slow breathing

(b

SRS (Affective)

407

habituation

group

0 Normal

(01

breathmg

group

(b) GAS

FIG. 2. -The effects of breathing manipulation upon the means and standard errors of the differences in the means of the SRS and GAS self-report scales: (a) prior to breathing manipulation; (b) immediately following manipulation; (c) after the recovery phase and prior to the habituation phase.

the comparison between the final minute of the recovery and pre-manipulation periods, ANOVA revealed only two significant effects. For NSCRs, the Minutes main effect was significant (F(1, 42) = 7.08, MSe = 7.77). Examination of the means displayed in Fig. 1 shows that this effect is accounted for by a decrease in NSCRs from the pre-manipulation to the recovery period. For RR, a significant interaction between Minutes and Breathing Manipulation was obtained (F(2, 42) = 5.01, MSe = 1.35). As can be seen in Fig. 1, this interaction is accounted for by an increase in RR for the SB group whereas the FB returned precisely to the pre-manipulation rate of 16.2 cycles per minute. In summary, these analyses indicate that the subjects had recovered either to or below pre-manipulation levels prior to the start of the habituation phase of the study.

Analysis of habituation phase measures The main purpose of the study was to demonstrate the differential effects of breathing rate on SCR magnitude. This was examined using a three-way ANOVA with Breathing Manipulation (3 levels), Stimulus Content (2 levels) and Trials (10 levels) as the main factors. Habituation was indicated by a significant Trials effect (Fl, 42) = 12.46, MSe = 0.00007). Although no other significant main effects or interactions were obtained, a number of significant trends were revealed. Significant linear components of the Trials X Breathing Manipulation (F(2, 14) = 8.67, MSe = 0.00005) and Trials X Stimulus Content (F(1, 7) = 10.28, MSe = 0.0003) were observed. These data are portrayed in Fig. 3 and indicate that the FB group resulted in larger SCR magnitudes which declined more steeply over trials. Similarly, the agoraphobic-relevant stimuli elicited larger initial SCRs but reached a lower asymptotic level over trials than the neutral stimuli. Habituation of SCRs was also assessed using changes in response probability over trials and the number of stimuli required before three consecutive zero responses were observed. Two-way ANOVAs were employed with Breathing Manipulation and Stimulus Content as the main factors. For both probability and

L. D. KARTSOUNIS and G. TLJRPIN

408

. Fast breathing group . Slow breathing group * Norma\ breathlnq group

FIG. 3.-Mean

skin conductance response (SCR) magnitude of breathing manipulation and trials.

as a function

trials-to-criterion measures, only the Breathing Manipulation factor was significant (F(2, 42) = 3.44 and 5.42, MSe = 0.08 and 12.25 respectively). Aplannedcomparison (F( 1, 42) = 6.08) indicated significantly greater response probabilities for the FB group (2 = 0.44) compared to the SB (2 = 0.18) and the NB (X = 0.27) control groups. Similarly, the planned comparison was also significant (F(l, 42) = 10.12) for trials-to-criterion. The means for the FB, SB and NB groups were 5.63, 1.69 and 2.75 respectively. It should be noted that there were no significant correlations between Locus of Control scores and any of the above measures of habituation. Tonic physiological measures were also examined in this phase but only revealed a significant Trials effect for SCL (F(1, 42) = 23.3, MSe = 0.01) which indicated a steady decrease in level throughout the habituation series. Similarly, there were no differences between groups on the SRS and GAS scores at the end of the habituation series. Finally, the post-experimental affective ratings of the photographs of the neutral and agoraphobic stimuli were analysed separately using two-way ANOVAs. There were no differences in the subjects’ ratings of the neutral stimuli. However, in the case of the agoraphobic photographs, there was some suggestion of differences in ratings between the FB, SB and NB groups (F(2, 42) = 2.56, MSe = 0.84, p < 0.10). The mean ratings for the FB, SB and NB groups were -0.56, -0.29 and +O. 16 respectively.

DISCUSSION

The main purpose of this study was to present an analogue demonstration of the effects of HV on the acquisition of fear responses to neutral and potentially

Hyperventilation

and electrodertnal

habituation

409

agoraphobic stimuli. It was predicted that overbreathing induced HV would lead to a negative affective state which in turn would result in the differential processing of contiguously presented visual stimuli as compared to stimuli presented in the context of either slow or normal breathing. Accordingly, re-presentation of these stimuli in a subsequent habituation series should result in SCRs with greater magnitudes, which ought to display slower habituation in the overbreathing induced HV group. In addition, it was predicted that subjects in this group would evaluate these stimuli more negatively than in the other two groups. Finally, it was proposed that these effects would be more prominent for agoraphobia-relevant as opposed to neutral visual stimuli. Manipulation of subjects’ breathing rates did not result in any significant differences in the overall magnitude of the SCRs to the visual stimuli presented within the habituation phase. However, examination of changes in SCR magnitude over trials did reveal that the FB group compared to the other two groups displayed larger initial responses and tended to remain at a higher level during the stimulus series. Moreover, the other two habituation measures, response probability and trials-to-criterion, showed slower habituation in the FB compared to the other two groups. Analysis of the subjects’ affective ratings of the photographs of the visual stimuli did not reveal any significant differences. However, there was a non-significant tendency for the FB group to negatively evaluate the photographs with agoraphobia-relevant content. The above results partially support the proposal that a HV state may lead to the acquisition of fear responses to neutral stimuli. In particular, the habituation measures would indicate that the FB manipulation resulted in differential processing of the visual stimuli. The nature of the underlying mechanism obviously requires further investigation. For example, delayed habituation may have arisen either due to an associative process whereby neutral stimuli acquire negative valence by being paired with the hyperventilation state or due to the effects of this state upon habituation per se. Nevertheless, we would argue that these data provide some experimental evidence in support of the notion that HV states may be a sufficient condition for the acquisition of anxiety/fear responses to neutral cues. Accordingly, the relationship between HV and either panic disorder or agoraphobia could be accounted for in terms of its effects on the development of anxiety responses. In the clinical context, the development of panic may depend on a number of HV-induced anxiety experiences. It could be argued that the present analogue study not only reflects the effects of HV on current fear acquisition but may also indicate the potential for future conditioning given a prolonged exposure to both the HV state and relevant environmental cues. For example, differences in SCR elicitation and habituation probably reflect differences in orienting, which in turn may determine the ease of subsequent conditioning [29]. It is interesting to note that the present study observed differences only in habituation and not response magnitude indices of orienting. However, Ohman and Bohlin [30] demonstrated that habituation rate rather than response magnitude was a better predictor of subsequent conditioning. These analogue data may suggest, therefore, that the FB group may be more prone to develop future anxiety responses. To test this speculative prediction, a future study would require a subsequent conditioning phase which would follow both the manipulation and habituation phases of the current study.

410

L. D. KARTSO~NIS and G. TURPIN

A number of alternative interpretations could be placed upon these data. First, it could be argued that the observed differences between the groups were due not to an induced state of HV but arose from the breathing manipulation per se. Hyperventilation should be defined in terms of lowered arterial pCO?. Since pCOz was not assessed in the present study, we cannot be certain that subjects in the FB group actually hyperventilated. However, they did display symptoms associated with HV, as measured by the SRS, and also exhibited elevated HR which has also been observed in response to overbreathing induced lowered&O2 in several studies [31]. A future study is required which employs direct measures of HV. Second, even if it is accepted that the subjects in group FB did hyperventilate, it could be argued that the subsequent effects were not related to HV but were the result of differences in either the cognitive or exertional demands placed upon the subjects. In terms of cognitive demands, an attempt was made to control this in the NB group where subjects were required to attend to paced instructions but were not required to make any paced response. Furthermore, a subjective measure of ‘cognitive arousal’, the GAS, was also employed but failed to discriminate between the groups. Future research is required to rule out the possible confounding effects of cognitive and exertional demands within this paradigm. Third, the effect of breathing manipulation on habituation might be explained solely in terms of differences in physiological arousal which persisted from the manipulation phase throughout the remainder of the experiment. Bohlin (321 has demonstrated that increased arousal leads to the retardation of habituation. This possibility was excluded in the present study by the inclusion of the recovery period. On the basis of pilot work, it was expected that following recovery all subjects should have returned to their pre-manipulation levels and that there should be no differences between the groups. The results of the recovery period analyses confirmed that the FB group did not display elevated arousal prior to the habituation series. Finally, two other aspects of the study require comment. First, an attempt was made to investigate the equipotentiality of agoraphobia-relevant and neutral cues by manipulating on a between subjects basis the contents of the visual stimuli. The inclusion of these sub-groups did not reveal any significant effects. However, the experimental design assumed that the agoraphobia-relevant material would be fearrelevant to the sample of subjects employed. Several studies have indicated that different populations are prone to different types of fears and phobia [33] and the British undergraduates, in particular, are especially liable to social fears [34]. It is possible, therefore, that the use of more appropriate stimulus material in this study would have resulted in significant effects. Second, the Locus of Control Questionnaire was employed in order to investigate its proposed relationship to the acquisition of agoraphobia. No correlations were observed between this and any other variable measured. In summary, the present analogue study has provided some tentative support for the purported role of HV in the aetiology of anxiety disorders. As we discussed earlier, several authors [5, 14, 151 have proposed that HV results both in the experience of anxiety symptoms and may also have a specific effect on the acquisition of anxiety disorders. However, these models are based either upon clinical observations or essentially correlational studies and do not necessarily address the problem

Hyperventilation

and electrodermal

of aetiology. It is proposed that the analogue provide an alternative approach for investigating anxiety related disorders.

habituation

411

paradigm investigated here may the relationship between HV and

Acknowledgement-This experiment was conducted in partial fulfilment Psychology in the Department of Psychology, Plymouth Polytechnic.

for the award

of M.Sc. Clinical

REFERENCES

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22.

23. 24. 25. 26. 27. 28.

BASS C and GARDNER W. Emotional influences on breathing and breathlessness. J Psychosom Res 1985; 29: 599609. SKARBEK A. The significance of varieties in breathing behaviour in speech and at rest. Act Psyc SC 1969; 45: 217-258. SKARBEK A. A psychophysiological study of breathing behaviour. Brir J Psych& 1970; 116: 637641. SUESS WM, ALEXANDER AB, SMITH DD, SWEENEY HW, MARION RJ. The effects of psychological stress on respiration: a preliminary study of anxiety and hyperventilation. Psychophysl 1980; 17: 535-540. LUM LC. Hyperventilation: the tip and the iceberg. J Psychosom Res 1975; 19: 375-383. LUM LC. Hyperventilation and anxiety state. JR Sot Med 1981; 74: 14. LEWIS BI. Hyperventilation syndrome: a clinical and physiologic evaluation. Calif Med 1959; 91: 121-126. WOLFE J. Psychotherapy by Reciprocal Inhibition. California: Stanford University Press, 1958. WOLFE J. The Practice of Behavior Theraov (2nd edn). New York: Perrramon Press. 1973. AVENI CA, CLJ~TER HS. Controlled rebreathing. Behav Engin 1977; 4117-21. VAN DEN HOUT MA, GRIEZ E. Panic symptoms after inhalation of carbon dioxide. Brit J Psychiat 1984; 144: 503-507. SNAITH RP. A clinical investigation of phobias. Brit J Psychiat 1968; 114: 673-697. LOWENSTEIN H. A clinical investigation of phobias. Brit J Psychiat 1968; 114: 1196. EMMELKAMP PMG. Phobic and Obsessive-Compulsive Disorders: Theory, Research and Practice. New York: Plenum Press, 1982. GARSSEN B, VAN VEENENDAAL W, BLOEMINK R. Agoraphobia and the hyperventilation syndrome. Beh Res Ther 1983; 21: 643-649. CLARK DM, SALKOVSKIS PM, CHALKLEY AJ. Respiratory control as a treatment for panic attacks. J Behav Ther Exp Psvchiat 1985; 16: 23-30. RAPEE RM. A case of panic disorder treated with breathing retraining. J Behav Ther Exp Psychiat 1985; 16: 6365. LEY R. Agoraphobia, the panic attack and the hyperventilation syndrome. Beh Res Ther 1985; 23: 79-81. OHMAN A, ERIKSON A, FREDRIKSON M, HUGDAHL K, OLOFSSON C. Habituation of the electrodermal orienting reaction to potentially phobic and supposedly neutral stimuli in normal human subjects. Biol Psycho1 1974; 2: 85-93. KARTSOUNIS LD, PICKERSGILL MJ. Orienting responses to stimuli others fear. BritJ Clin Psychol 1981; 20: 261-273. MARTIN I, LEVEY AB. Evaluative conditioning. Adv Behav Res Ther 1978; 1: 57-102. OHMAN A. The orienting response, attention and learning: an information-processing perspective. In The Orienting ReJiex in Humans (Edited by KIMMEL HD, VAN OLST EH, ORLEBEKE JF). Hillsdale: Lawrence Erlbaum Associates, 1979. MATHEWS AM, GELDER MG, JOHNSTONE DW. Agoraphobia London: Tavistock Publications, 1981. ROTTER JB. Generalized expectancies for internal versus external control of reinforcement. Psycho/ Monogr 1966; 80: (whole No. 609). LYKKEN DT, VENABLES PH. Direct measurement of skin conductance: a proposal for standardisation. Psychophysl 1971; 8: 656672. CLARK DM, HEMSLEY DR. The effects of hyperventilation: individual variability and its relation to personality. J Behav Ther Exp Psychiat 1982; 13: 4147. CLEMENTS PR, HAFER MD, VERMILLION ME. Psychometric, diurnal and electrophysiological correlates of activation. J Pers Sot Psycho! 1976; 33: 387-394. GREENHOUSE SW, GEISSER S. On methods in the analysis of profile data. Psychometrika 1959; 24: 95-l 12.

412 29. 30. 31.

32. 33.

34.

L. D. KARTSOUNIS

and G. TURPIN

OHMAN A. The orienting response during Pavlovian conditioning. In Orienting and Habituarwn: Perspectives in Human Research (Edited by SIDDLE DAT). Chichester: Wiley, 1983. OHMAN A. and BOHLIN G. Magnitude and habituation of the orienting reaction as predictors of discriminative electrodermal conditioning. .I Exp Res Pers 1973; 6: 293-299. GROSSMAN P, WIENTJES K. Respiratory-cardiac coordination as an index of cardiac functioning. In Psychophysiology of Cardiovascular Control: Models, Methods, and Data (Edited by ORLEHEKE JF, MULDER G, VAN DOORNEN LJP). New York: Plenum Press, 1985. BOHLIN G. Delayed habituation of the electrodermal response as a function of increased level of arousal. Psychophysl 1976: 13: 345-351. WI~TKOWER ED. and DURREUIL G. Reflections on the interface between psychiatry and anthropology. In The Interface Between Psychiatry and Anthropology (Edited by GAI.DSTON I). New York: BrunueVMazel, 1971. KARTSOUNIS LD, MERVYN-SMITH J, PICKERSGILL MJ. Factor analysis of the responses of British university students to the Fear Survey Schedule (FSS-III). Person Indiv Diff 1983: 4: 157-163.