Panic disorder: the role of the balance system

Panic disorder: the role of the balance system

Giampaolo Pernaa,*, Alpini Dariob, Daniela Caldirolaa,Barozzi Stefaniac, Antonio Cesaranid, Laura Bellodia

aAnxiety Disorder Clinical and Research Unit, Department of Neuropsychiatric Sciences, Vita-Salute University,

Istituto Scientifico HS Raffaele, 29 via Prinetti 20127, Milan, ItalybScientific Institute S. Maria Nascente don Gnocchi Foundation, via Capecelatro 66, Milan, Italy

cInstitute of Otolaryngology, University of Milan, via F. Sforza 35, Milan, ItalydInstitute of Otolaryngology, University of Sassari, via Mancini 5, Sassari, Italy

Received 10 October 2000; received in revised form 5 June 2001; accepted 13 July 2001

Abstract

Experimental evidences suggest that Panic Disorder (PD) is characterized by abnormalities in respiratory and vestibular func-tions. We studied balance system function in patients with PD and its relationships with CO2 reactivity and clinical characteristics.

Nineteen patients with PD with/without agoraphobia underwent static posturography and the 35% CO2 challenge. The severity ofclinical symptomatology was measured by standardized psychometric scales. Patients were free of psychotropic medications duringthe 2 weeks before the study. Different investigators blind to each other carried out the CO2 challenge, static posturography and

clinical assessment. Nineteen age and sex-matched healthy controls underwent static posturography. Body sway velocity and lengthwere significantly higher in panic patients than in controls and patients showed high percentages of abnormal scores. Patients withtwo or more abnormal scores on static posturography were significantly more agoraphobic than those with less than two. Abnor-mal posturography scores under the eyes-opened was related to high anticipatory anxiety, whereas those under eyes-closed was

related to phobic avoidance. Symptomatological reactivity to CO2 was significantly correlated to abnormal functions of the balancesystem in the eyes-closed condition. Our findings suggest that (1) many patients with PD (5–42%) have abnormalities in their bal-ance system function compared with healthy controls (0–5%), (2) symptomatological reactivity to CO2 and balance system function

in patients with PD are correlated only in the eyes-closed condition and (3) there is a significant link between agoraphobic avoid-ance and subclinical abnormal function of the balance system network. # 2001 Elsevier Science Ltd. All rights reserved.

Keywords: PD; Respiration; Balance system; Agoraphobia; Panic provocation; Anxiety; Brainstem

1. Introduction

Significant relationships between balance functiondisorders and panic disorder (PD) has been repeatedlyreported (Jacob, 1988; Jacob et al., 1989, 1992; Sklare etal., 1990; Clark et al., 1994; Frommberger et al., 1994;Hoffman et al., 1994; Stein et al., 1994; Yardley et al.,1994, 1995; Furman and Jacob, 1997; Gordon et al.,1998; Asmundson et al., 1998). The studies into theserelationships have focused on two main areas: vestibu-lar abnormalities in patients with PD, and anxietysymptoms or disorders among patients with abnormal-

ities in the vestibular system. Jacob and co-workers(Jacob et al., 1985, 1989) reported vestibular abnormal-ities in patients with PD who complained of dizziness.This finding has been replicated in patients with PDnon-specifically selected for the presence of dizziness(Sklare et al., 1990; Hoffman et al., 1994; Jacob et al.,1997), where the abnormalities were not the expressionof a full-blown vestibular disorder. Similarly, patientswith vestibular disorders have psychiatric syndromeswith a particular reference to PD and agoraphobia(Hallam and Stephens, 1985).Yardley and co-workers (1995) reported an associa-

tion between vestibular abnormalities and phobicavoidance. Jacob and co-workers (1996) showed thatpatients with PD and agoraphobia had more vestibularabnormalities than patients with PD but without

0022-3956/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.

PI I : S0022-3956(01 )00031-0

Journal of Psychiatric Research 35 (2001) 279–286

www.elsevier.com/locate/jpsychires

* Corresponding author. Tel.: +39-02-26433211; fax: +39-02-

26433265.

E-mail address: [email protected] (G. Perna).

agoraphobia, patients with depressive disorders or withanxiety disorders other than panic or healthy controls.We (Allevi et al., 1997) found a significant relationshipbetween the presence of dizziness during panic attacks andagoraphobia. It seems that in some patients phobicavoidance might not be a mere consequence of panicattacks but rather the result of a more complex interaction.Abnormalities in the respiratory function may play a

central role in the pathogenetic mechanisms underlyingPD and has been mainly associated with panic attacks(Klein, 1993; Bellodi and Perna, 1998). Some observa-tions have indicated that the vestibular and the respira-tory systems may work in complex interrelationshipswith one another. Dizziness is one of the main symp-toms induced by CO2 inhalations (Perna et al., 1994a)and by hyperventilation (Fried and Grimaldi, 1993).Balance and respiratory systems seem to influence two

different aspects of panic disorder, respectively agor-aphobia and panic attacks. Control centers of both,however, are located nearby in the brainstem with sig-nificant reciprocal influences and there is evidence ofvestibular-respiratory functional interconnections (Fur-man et al., 1998). We hypothesize that balance functionmight play a role in the development of agoraphobiaand that respiratory reactivity might influence this pro-cess. To test this hypothesis we evaluated the balance sys-tem function by posturography and respiratory reactivityby CO2 inhalation test in a sample of patients with PD.

2. Methods and materials

2.1. Subjects

Nineteen patients with PD with/without agoraphobiawere included in the study [13 women and six men,mean age 27.1�9.1 years (17–46 years), age at onset:20.1�4.8]. The patients were consecutively recruitedover 8 months at the outpatient facilities of the AnxietyDisorders Clinical and Research Unit at San RaffaeleHospital in Milan. Eighteen patients (95%) admitted tothe study had a lifetime history of agoraphobia. Sinceno normative data for the Italian general population onstatic posturography are available in international peer-reviewed journals, a sample of 19 sex- and age-matchedhealthy controls were recruited by advertisementsplaced around the University and tested with staticposturography. Controls had never fit any lifetime psy-chiatric diagnoses and had never experienced panicattacks. Consensus diagnoses according to DSM IVcriteria were established by two senior psychiatrists whoindependently assessed patients by a clinical interviewand the MINI International Neuropsychiatric Inter-view-Plus (Sheehan et al., 1994). Exclusion criteria forall the patients were the presence of psychiatric dis-orders other than PD with/without agoraphobia.

Both patients and controls underwent a physical andneuro-ophthalmologic examinations. Accurate medicalhistory taking were performed to exclude any subjectswith substantial current medical disorders, especiallyrespiratory and balance systems disorders, neurologicaldeficits, signs of polyneuropathies, personal and familialhistory of cerebral aneurysm, significant hypertension(systolic > 180 mm Hg, diastolic > 100 mm Hg),pregnancy or epilepsy. No subject included in the studyhad been treated with ototoxic drugs such as aminogly-cosides. The procedure of this study has been carriedout in accordance with the Declaration of Helsinki andall subjects gave their informed consent to the studyafter receiving a detailed explanation of the entire pro-cedure. During the 2 weeks before the start of the study,all subjects were free of psychotropic medications. Theywere asked to refrain from alcohol and any kind ofmedication for at least 2 days, beverages containingxanthines for at least 12 hours and food or smoking forat least 2 hours before both static posturography andCO2 challenge.

2.2. General procedure

The patients underwent a series of tests according tothe following schedule:

Day 0: psychometric tests and interviews for diag-nostic evaluation and clinical assessment.Day 1: static posturography.Day 3: the 35% CO2 challenge.

2.2.1. Clinical assessmentThe severity of PD was evaluated by the Panic Asso-

ciated Symptoms Scale (PASS) to obtain a global score(PASS-tot) and subscores for panic attacks, phobicavoidance and anticipatory anxiety (Argyle et al., 1991).Patients also completed the following self-administeredscales:

. The Fear Questionnaire (FQ) to obtain a globalscore (FQ-tot) and subscores for agoraphobia(FQ-AGO), blood-injury phobia (FQ-BI) andsocial phobia (FQ-SP; Marks and Mathews, 1979).

. The Sheehan Disability Scale (SDS) to obtain aglobal score (Leon et al., 1992).

. The Dizziness Handicap Scale (DHI) to obtain aglobal score (Jacobson and Newman, 1990).

2.2.2. Static posturography (stabilometry)In general, static posturography gives information on

the ability to integrate the multiple inputs that con-tribute to the control of posture. Although posturo-graphy is not able to give any specific neuroanatomicalinformation it can point out the need for re-educationof the sensory integration when the system does not

280 G. Perna et al. / Journal of Psychiatric Research 35 (2001) 279–286

work properly (Gagey, 1991) and it can explain residualdizziness-related symptoms in subjects without clinicallydefinite balance disorders.Patients were recorded, between 4 and 6.00 p.m., on a

standard vestibular platform conforming to the stan-dards of the International Society of Posturology (Bizzoet al., 1985), one of the two recognized internationalsocieties dedicated to posture (the other one is theInternational Society of Posture and Gait). This mova-ble platform is used to destabilize subjects and allowsthe evaluation of postural strategy and tonic posturalactivity by pressure sensors located in the platform.From the signals generated by the sensors, the changesin successive positions of the center-of-pressure can becalculated to create the area of a statokinesigram, whichdescribes how accurately a subject can keep steadywithin his/her environment. All the patients were recor-ded in the eyes-opened, eye-closed and neck-extensionconditions. For each condition, the following para-meters characterizing body sway (Aalto et al., 1988;Gagey, 1991) were recorded:

. the surface of the body sway (Area), expressed inmm2, represented by the confidence ellipse con-taining 90% of the sample position indicating theprecision of the system;

. the length of the body sway (Length), expressed inmillimeters, represented by to total distance fromthe center of pressure of the subject which indicatethe energy spent; and

. the mean velocity of body sway (Velocity),expressed in mm/s, again expression of the energyspent.

2.2.3. Thirty five percent CO2 challenge2.2.3.1. Apparatus. Two different gas mixtures wereused: compressed air (placebo) and a mixture of 35%CO2 and 65% O2. Both gases were inhaled through thesame self-administration mask. Vital capacity was mea-sured on a respirometer (Wright Respirometer Mark 20,Ferraris Medical Ltd) connected to the self-administra-tion mask. The same respirometer measured the gasvolume delivered in each inhalation.

2.2.3.2. Procedure. All subjects were tested, between 4and 6.00 p.m., in a double-blind, random, cross-overdesign, according to the method described by Griez andco-workers (1990) They were informed they would beinhaling two harmless gas mixtures containing differentpercentages of CO2 and O2, and that they mightexperience some discomfort, felt as a few neurovegeta-tive symptoms or a definite sensation of anxiety/dis-comfort with several somatic and/or cognitivesensations. But the word ‘‘panic attack’’ was not men-tioned. Vital Capacity was measured and baseline anxi-ety assessed by the State-Trait Anxiety Inventory for

state anxiety (STAI; Spielberger et al., 1970). Each sub-ject then inhaled one vital capacity of 35% CO2/65% O2

or of compressed air in randomly assigned order at aninterval of 25 to 30 min. After each inhalation the sub-jects were asked to hold their breath for 4 s. The testwas considered valid only if the subject had inhaled atleast 80% of the measured vital capacity. Immediatelybefore and after each inhalation (air or CO2) anxietywas evaluated on two scales. The Panic Symptom List(PSL III-R; Pols et al., 1991) is a self-rating ques-tionnaire that assesses the 13 panic symptoms describedin DSM III-R/IV on a five point scale (0=absent,1=mild, 2=moderate, 3=severe, 4=very intense) toobtain a Total Symptom Score (TSS) (range 0–52). TheVisual Analogue Scale for Anxiety (VAS-A) describesthe degree of global subjective anxiety on a continuumfrom zero (no anxiety present) to 100 (the worst anxietyimaginable). The Human Ethics Committee of SanRaffaele Hospital approved the procedure of the 35%CO2 test.

2.2.3.3. Quantitative assessment. The anxiety reactivityto 35% CO2 inhalations was evaluated as �% VAS-A(the percentage of maximum increment or decrementpossible on the VAS-A scale; Perna et al., 1994b), andcalculated as follows:

1. If � VAS-A (post-CO2 VAS-A values minus pre-CO2 VAS-A values) was positive, then �% VAS-A=�VAS-A�100/(100�VAS-A before CO2).

2. If �VAS-A was negative, then �% VAS-A=�VAS-A�100/VAS-A before CO2.

The symptomatological reaction to CO2 was eval-uated as �TSS (TSS post inhalation–TSS before inha-lation).

2.2.3.4. Qualitative assessment. According to the idealthreshold obtained by Receiver Operating Character-istic (ROC) analysis of the 35% CO2 challenge (Batta-glia and Perna, 1995), the reaction was considered‘‘positive’’ if �% VAS-A 526 and ‘‘negative’’ if �%VAS-A <26.

2.3. Data analyses

Static posturography scores were subdivided intonormal or abnormal values. The scores were consideredabnormal if they exceeded the mean values+2 S.D.s ofscores obtained in the sample of healthy controls exam-ined. With the aim to look for differences related to thepresence of few or many posturographic abnormalities,the patient sample was also subdivided in two groups, aHigh Abnormal (HA, n=9, age: 26.3�10.3 years, 3/9males) and a Low Abnormal (LA, n=10, age: 27.8�8.3years, 3/10 males) group, using as the cut-off score themedian number of abnormalities (n=2) from the static

G. Perna et al. / Journal of Psychiatric Research 35 (2001) 279–286 281

posturography test. The median strategy was chosen toobtain two subgroups with similar number of patients.The patients were also regrouped according to the pre-sence or absence of abnormal posturographic scores foreach testing condition.The significance of any difference in continuously

distributed variables between two groups was examinedby analyses of variance (ANOVA) and multivariateanalyses of variance (MANOVA) with post-hoc New-man–Keuls Tests. Separate MANOVAs were used whencomparing psychometric, posturographic and CO2

reactivity scores. Chi-square analyses, Spearman RankOrder correlation or Fisher tests were applied to assessdifferences between groups for nominal values. Fisher’sexact tests with Bonferroni’s correction were applied tocompare the pattern of distribution of subjects withabnormal posturographic scores between patients andcontrols. Spearman Rank Correlation tests with Bon-ferroni’s correction were applied to evaluate the corre-lations between reactivity to 35% CO2 and staticposturographic measures.

3. Results

The clinical characteristics of the sample and CO2

reactivity are listed in Table 1. The scores on the psy-chometric scales and CO2 reactivity tests equaled thoseof other patient populations with PD but were higherthan those of healthy controls (Perna et al., 1994a,1995). The DHI scale score was higher than those of thegeneral population (Guidetti, 1989) and similar to thoseof patients with peripheral vestibular system disorders(Mann et al., 1996).MANOVA showed significant differences in static

posturography measures between patients with PD andhealthy controls (Wilks’ Lambda: 0.48, Rao’s R: 3.8,d.f.: 9.28, P<0.007). Post-hoc Newman–Keuls testsshowed significantly higher scores in patients with PDfor Velocity parameter in all the conditions and Lengthparameter in the eyes-opened and neck-extension con-ditions (Table 2). Patients with PD had high percentagesof abnormal scores on static posturographic parameterssignificantly higher than controls for Length and Velo-city parameters in eyes-closed and neck-extension con-ditions (Table 2). More than one third of the patientstested showed abnormal scores for Length and Velocityparameters. Twelve patients (63%) had at least oneabnormal static posturography score and nine (47%)had two or more.Assessing PASS-global score, patients in the HA

group (10.6�3.1) were significantly (F=5.2, d.f.=17,P<0.05) more severely ill than those in the LA group(6.7�4.2). MANOVAs showed significantly differencesin psychometric subscales scores between the HA andthe LA groups (Wilks’ Lambda: 0.35, Rao’s R: 3.8, d.f.:

6.12, P<0.05). Post-hoc Newman–Keuls tests showedsignificantly (P<0.04) higher scores for PASS-avoid-ance in the HA than in the LA group (1.7�0.9 vs.0.7�0.9). Compared with patients who achieved nor-mal posturographic scores in the eyes-closed condition,the patients with abnormal scores showed more severephobic symptomatology (Wilks’ Lambda: 0.32, Rao’sR: 4.3, d.f.: 6.12, P<0.02; see Table 3) while in the eyes-opened condition, those with abnormal scores showedhigher scores for PASS-anticipatory anxiety (Wilks’Lambda: 0.37, Rao’s R: 3.4, d.f.: 6.12, P<0.04, post-hoc Newman–Keuls test, P<0.03). Table 4 reports therates of agoraphobia according to normality/abnorm-ality of posturographic tests. The patients in the HAgroup showed a more severe agoraphobia than those inthe LA group. This difference was mainly related to thetests made in the eyes-closed condition. In patients withPD there were no significant correlations between psy-chometric scale scores and posturographic measures.No significant differences for measures of CO2 reac-

tivity (�% VAS-A, �TSS and number of symptomsincrease after CO2) and for patterns of distribution ofpositive and negative reactors to CO2 were found com-paring HA and LA groups or comparing patients withnormal and abnormal scores on any of the three pos-turographic testing conditions. Table 5 reports the cor-relations between measures of CO2 reactivity andposturographic scores. The symptomatological reactionto CO2 correlated significantly with static posturo-graphy scores obtained in eyes-closed condition whileno correlation was found between �% VAS-A and anyof the posturographic measures.

Table 1

Clinical characteristics and CO2 reactivity of patient with PD

Mean�S.D.

Clinical characteristics

STAI 52.3�12.8

Dizziness Handicap Inventory-Global Score 38.0�19.3

Fear Questionnaire-Global Score 45.3�25.7

Sheehan Disability Scale-Global Score 15.4�7.4

Panic Associated Symptom Scale:

Global score 8.5�4.1

Panic attacks score 4.2�2.7

Anticipatory anxiety score 3.2�1.8

Avoidance score 1.2�1.1

CO2 reactivity

VAS-A before CO2 31.9�23.4

VAS-A after CO2 65.7�29.8

�% VAS-A 54.0�33.8

TSS before CO2 6.4�7.1

TSS after CO2 19.2�10.7

� TSS 12.8�9.0

Increase in number of symptoms after CO2 6.4�3.0

VAS-A, Visual Analogue Scale for Anxiety; TSS, Total Symptom

Scale.


No significant correlations were found between thescores for agoraphobic avoidance (FQ-AGO and PASS-Avoidance) and CO2 reactivity (�% VAS-A, post-CO2

VAS-A, � TSS and post-CO2 TSS), nor was there anysignificantly different distribution pattern of positiveand negative responses between patients with no/lowagoraphobia and those with moderate/severe agor-aphobia.

4. Discussion

We have studied the balance system function in rela-tion to respiratory abnormal reactivity in patients with

PD. The results should be considered preliminary fortwo reasons. The sample size could create type II errorsand the absence of comparison groups with other psy-chiatric disorders precludes discussion on the specificityof our findings. Moreover, static posturography mea-sures postural sway, a valid indicator of the balancesystem function. However, this technique is relativelyunspecific since it is unable to distinguish between cen-tral or peripheral nervous system disorders of the bal-ance system and there is a debate on its ability todistinguish simulated disorders (Uimonen et al., 1995).

Table 2

Static posturography in patients with PD and controls

Healthy controls Patients with PD

Mean�S.D. No. (%) of abnormal scores Mean�S.D. No. (%) of abnormal scores

Eyes-open condition

Area 210�78.0 0 200�122 1 (5%)

Lengtha 340�43 1 (5%) 427�124 7 (37%)

Velocitya 6.7�0.8 1 (5%) 8.4�2.5 7 (37%)

Eyes-closed condition

Area 304�97 0 315�169 4 (21%)

Lengthb 545�75 0 658�256 7 (37%)

Velocitya,b 10.0�1.5 0 13.0�5.0 8 (42%)

Neck extension condition

Area 377�102 1 (5%) 400�198 2 (11%)

Lengtha,b 609�75 0 747�268 7 (37%)

Velocitya,b 12.1�1.4 0 14.7�5.3 7 (37%)

No. of abnormal scores 0.2�0.4 2.6�2.6

Area, mm2; Length, mm; Velocity, mm/sec.a MANOVA, post-hoc Newman–Keuls tests: P<0.05.b Fisher exact test with Bonferroni’s correction: P<0.05.

Table 3

Severity of panic–phobic symptomatology in patients with or without

static posturographic abnormalities in eyes-closed condition

Normal

(n=11)aAbnormal

(n=8)b

Fear questionnaire

Agoraphobiac 10.1�11.6 23.0�9.9

Blood Injury Phobia 19.5�14.1 19.6�9.1

Social Phobiac 6.2�7.9 15.8�5.9

Panic Associated Symptom Scale

Panic attacks score 3.4�2.7 5.3�2.3

Anticipatory anxiety score 2.5�1.9 4.1�1.5

Avoidance scorec 0.6�0.9 1.9�0.6

a Normal: with all the posturographic parameters (Area, Length,

Velocity) within the normal range.b Abnormal: at least one posturographic parameter higher than the

mean+2 S.D. of scores obtained among controls.c MANOVA, post-hoc Newman–Keuls tests: P<0.05.

Table 4

Posturographic abnormalities and agoraphobic avoidance in patients

with PD

Agoraphobic avoidance

None/little Moderate/severe

Posturography scores in eyes-opened condition:

All normal 9 (75%) 3 (25%)

One or more abnormal 3 (43%) 4 (57%)

Posturography scores in eyes-closed conditiona:

All normal 10 (91%) 1 (9%)


Posturography scores in neck-extension condition:

All normal 9 (75%) 2 (25%)


Posturography scores in all conditionsb:

Less that two abnormal measures 9 (90%) 1 (10%)

Two or more abnormal measures 3 (33%) 6 (66%)

a Fisher test: P<0.007, Spearman Rank Test: t=3.8, P<0.002.b Fisher test: P<0.02, Spearman Rank Test: t=3.0, P<0.009.


Despite these limitations, our findings indicate that(1) many patients with PD (5–42%) have abnormalitiesin their balance system function compared with healthycontrols (0–5%), (2) symptomatological reactivity toCO2 and balance system function in patients with PDare correlated only in the eyes-closed condition and (3)there is a significant link between the severity of panicdisorder, in particular agoraphobic avoidance, andsubclinical abnormal function of the balance systemnetwork.Abnormal scores on balance system function tests

appear frequently among patients with PD withoutclinical vestibular disorders. Abnormalities in the bal-ance system function are related to illness severity andour data confirm the strong association between balanceabnormalities and agoraphobic avoidance (Yardley etal., 1995; Jacob et al., 1996), but no significant rela-tionships were found with panic attacks. Nine (90%) ofthe 10 patients with less than two abnormal scores onstatic posturography reported no or little agoraphobia,whereas six (67%) of the nine patients with two or moreabnormal scores reported moderate to high agor-aphobia. This difference was only detectable in the eyes-closed condition, while in the eyes-opened conditionpatients with abnormal posturographic scores showedmore severe anticipatory anxiety. These differencespoint to the role visual information plays in thesepatients compensating for the balance deficits possiblyrelated to agoraphobic avoidance. But with an increasedawareness of the environment achieved in the eyes-opened condition, able to stabilize balance systeminstability, greater anticipatory anxiety might alsoresult.

Among panic patients, symptomatological reaction toCO2 correlated with the area of body sway, indicates theprecision of the balance system. This relationship couldbe explained by the integration of the neuronal circuitsin the brainstem. The locus coerules, the raphe nucleiand the ventrolateral medulla are the main brainstemnuclei. These areas are thought to be important in theetiopathogenesis of PD (Bellodi and Perna, 1998). Theyform part of a broad brainstem respiratory network(Nattie, 1999) and the vestibular nuclei influence andare influenced by all of them (Furman et al., 1998;Yates, 1998; Yates and Stocker, 1998). The highlyinterconnected vestibular and respiratory systems(Yates et al., 1993; Miller et al., 1995; Miller andSiniaia, 1996; Balaban, 1996; Yardley et al., 1998;Balaban and Thayer, 2001) maintain blood gas home-ostasis during movement and changes in posture. Asubtle abnormal regulation of the vestibular systemcould trigger abnormal respiratory sensations able totrigger fear and avoidance symptomatology in reactionto motion or space discomfort. Moreover, mental dis-tress in phobic situations may increase arousal andhyperventilation, both conditions that can furtherdestabilize the balance system, thus leading to a viciouscircle of distress and destabilization. Finally, the factthat the symptomatological reaction to CO2 correlatedwith static posturographic scores only in the eyes-closedbut not in the eyes-opened and neck-extension condi-tions, support again the idea that visual informationmight be important for compensating the influences ofthe balance system abnormalities on PD vulnerability.The absence of measures of CO2 reactivity among heal-thy controls, however, does not allow to make infer-ences on the specificity of the relationship of CO2

reactivity and static posturographic abnormalities inpanic disorder.Balance control system is based on multiple and

redundant sensory information. The system maintains anormal balance even when one of its sub-systems isdamaged or dysfunctional because central compensa-tion and reorganization can take place. In the globalsensory strategy to regulate the balance system, Jacob etal. (1995) have suggested that visual information is nee-ded to compensate the subclinical abnormalities in thebalance system of patients with PD. The importance ofvisual information in compensating the balance systemnetwork is confirmed by our results showing that inagoraphobic patients balance system reacts abnormallymainly in the eyes-closed condition, when visual infor-mation is lacking. In daily life situations with ‘‘particu-larly rich and repetitive visual patterns’’ (Bronstein,1995; crowds, open or closed spaces, traffic), the visualinformation might be inadequate to stabilize the dys-functional balance system, so that this lack could triggerdizziness that in turn could strengthen the tendency tophobic avoidance.

Table 5

Correlations between static posturography and CO2 reactivity scores

in patients with PD

�%

VAS-A

�

TSS

Number of symptoms

increased after CO2

Eyes-opened condition

Area 0.26 0.40 0.29

Length 0.35 0.31 0.26

Velocity 0.35 0.39 0.24

Eyes-closed condition

Area 0.33 0.67a 0.68a

Length 0.36 0.59 0.54

Velocity 0.36 0.59 0.54

Neck-extension condition

Area 0.15 0.50 0.44

Length 0.19 0.48 0.36

Velocity 0.19 0.48 0.36

�% VAS-A, global anxiety reactivity to CO2; �TSS, global sympto-

matological reactivity to CO2.a Spearman Rank Test: P<0.0016 (level of significance after Bon-

ferroni’s correction P<0.0019).


A balance system network dysfunction might be oneof the components of the psychobiological mechanismsleading to agoraphobia. The latter might be the result ofcomplex interactions between higher and lower brainfunctions. Agoraphobic avoidance is conceptualized asthe result of contextual learning, i.e. association ofpanic attacks with the context in which they occurred.This process involves mainly the medial pre-frontalcortex and the limbic systems (Gorman et al., 2000),however, the association between balance function andagoraphobia suggest that inputs from the balance sys-tem network could represent sub-clinical reinforcementsable to influence strongly the higher brain mechanismsinvolved in the development of agoraphobia.An alternative explanation to the existence of a bal-

ance system network dysfunction is that the abnormalposturographic tests might reflect a change in posturalstrategy, as an adaptation process to panic–phobicsymptomatology, rather than a sensori-motor dysfunc-tion of the balance system. The latter idea is based onrecent studies in patients with ‘‘phobic postural ver-tigo’’, a disorder characterized by persistent dizziness,subjective disturbances of balance, attacks of posturalinstability with/without anxiety and avoidance beha-viors in the absence of objective pathology of the bal-ance system. In these studies (Krafczyk et al., 1999;Querner et al., 2000) the authors observed an increase inbody sway during normal upright stance in the higherfrequency band with a disappearance of the differencefrom controls during more difficult balance tasks. Theyinterpreted these findings as an expression of the con-scious unnecessary application of a postural strategy ofcontraction of the anti-gravity muscles as a balancestrategy of caution. The latter explanation, however,appears to be less plausible to our results since, differ-ently from patients with phobic postural vertigo, per-sistent dizziness and postural instability are not themain clinical aspects of patients with PD and, in parti-cular, those included in our study experienced dizzinessonly during panic attacks and thus conscious control ofstance is unlikely. Moreover, phobic avoidance inpatients with phobic postural vertigo and in patientswith PD is the result of different fears, the first related tothe consequences of postural dizziness (e.g. fallingdown) and the second to the occurrence of panicattacks. Only in the first case, balance strategy of cau-tion is likely.Anyway, the experience of dizziness could induce fear

and avoidant behaviors through the activation of theamygdala, the limbic system and the cerebral cortexcircuitry (Costin et al., 1970; Guldin and Grusser, 1998).On the one hand, vestibular inputs to the limbic systemand the amygdala may be thought of as a neuroanato-mical pathway of the balance systems vulnerability toagoraphobia. On the other hand, vestibular abnormal-ities may be considered a secondary consequence of

panic attacks or anxiety. In turn, anxiety-inducedabnormal function of the balance system could lead tothe development of agoraphobia.Overall, our results underline the role the balance

system has in the behavioral abnormalities of PD.Although it seems unlikely that an abnormal function ofthis system could ultimately cause PD, the balance sys-tem appears to influence the psychobiological mechan-isms underlying agoraphobic avoidance.

References

Aalto H, Pyykko I, Starci J. Computerized posturography, a devel-

opment of the measuring system. Acta Otolaryngology (Stockh)

1988;449:71–5.

Allevi L, Perna G, Bussi R, Bertani A, Bellodi L. Dizziness, panic

attacks and agoraphobia. European Neuropsychopharmacology

1997;7:s230.

Argyle N, Deltito J, Allerup P, Albus M, Nutziger D, Rasmussen S.

The Panic-Associated Symptoms Scale: measuring the severity of

PD. Acta Psychiatrica Scandinavica 1991;83:20–6.

Asmundson GJ, Larsen DK, Stein MB. PD and vestibular dis-

turbance: an overview of empirical findings and clinical implica-

tions. Journal of Psychosomatic Research 1998;44:107–20.

Balaban CD. Vestibular nucleus projections to the parabrachial

nucleus in rabbits: implications for vestibular influences on the

autonomic nervous system. Experimental Brain Research 1996;108:

367–81.

Balaban CD, Thayer JF. Neurological bases for balance-anxiety links.

Journal of Anxiety Disorders 2001;15:53–79.

Battaglia M, Perna G. The 35% challenge CO2 in PD: optimization by

Receiver Operating Characteristic (ROC) analysis. Journal of Psy-

chiatry Research 1995;2:111–9.

Bellodi L, Perna G. The panic respiration connection. Milan: MDM

Medical Media, 1998.

Bizzo G, Guillet N, Patatt A, Gagey PM. Specification for building a

vertical force platform designed for clinical stabilometry. Medicine

and Biology Engineer Computing 1985;23:474–6.

Bronstein AM. Visual vertigo syndrome: clinical and posturography

findings. Journal of Neurology and Neurosurgery 1995;59:472–6.

Clark DB, Hirsch BE, Smith MG, Furman JMR, Jacob RG. Panic in

otolaryngology patients presenting with dizziness or hearing loss.

American Journal of Psychiatry 1994;151:1223–5.

Costin A, Hafemann D, Elazar Z, Adey WR. Posture and the role of

vestibular and proprioceptive influences on neocortical, limbic,

subcortical and cerebellar EEG activity. Brain Research 1970;17:

259–75.

Fried R, Grimaldi J. The psychology and physiology of breathing. In:

Behavioral medicine, clinical psychology and psychiatry. New York

and London: Plenum Press, 1993.

Frommberger UH, Tettenborn B, Buller R, Benkert O. PD in patients

with dizziness. Archives of Internal Medicine 1994;154:590–1.

Furman JM, Jacob RG. Psychiatric dizziness. Neurology 1997;48:

1161–6.

Furman JM, Jacob RG, Redfern MS. Clinical evidence that the ves-

tibular system participates in autonomic control. Journal of Vestib-

ular Research 1998;1:27–34.

Gagey PM. Non vestibular dizziness and static posturography. Acta

Otorhinolaryngoiatrica Belgica 1991;45:335–9.

Gordon JG, Asmundson GJG, Larsen DK, Stein MB. PD and ves-

tibular disturbance: an overview of empirical findings and clinical

implications. Journal of Psychosomatic Research 1998;1:107–20.


Gorman JM, Kent JM, Sullivan GM, Coplan JD. Neuroanatomical

hypothesis of PD, revised. American Journal of Psychiatry 2000;

157:493–505.

Griez E, de Loof C, Pols H, Zandbergen J, Lousberg H. Specific sen-

sitivity of patients with panic attacks to carbon dioxide inhalation.

Psychiatry Research 1990;31:193–9.

Guidetti G. Stabilometria clinica. Milan: Amplifon Edizioni, 1989.

Guldin WO, Grusser OJ. Is there a vestibular cortex? Trends in Neu-

rological Sciences 1998;21:254–9.

Hallam RS, Stephens SDG. Vestibular disorder and emotional dis-

tress. Journal of Psychosomatic Research 1985;29:407–13.

HoffmanDL, O’Leary DP,Munjack DJ. Autorotation test abnormalities

of the horizontal and vertical vestibulo-ocular reflexes in PD. Archives

of Otolaryngology, Head and Neck Surgery 1994;110:259–69.

Jacob RG, Moller MB, Turner SM, Wall III C. Otoneurological

examination of PD and agoraphobia with panic attacks: a pilot

study. American Journal of Psychiatry 1985;142:715–20.

Jacob RG. PD and the vestibular system. Psychiatric Clinic of North

America 1988;11:361–74.

Jacob RG, Lilienfeld SO, Furman JM, Turner SM. Space and motion

phobia in PD with vestibular dysfunction. Journal of Anxiety Dis-

order 1989;3:117–30.

Jacob RG, Furman JM, Clark DB, Durrant JD. Vestibular symptoms,

panic and phobia: overlap and possible relationships. Annals of

Clinical Psychiatry 1992;4:163–74.

Jacob RG, Redfern MS, Furman JM. Optic flow-induced sway in

anxiety disorders associated with space and motion discomfort.

Journal of Anxiety Disorder 1995;9:411–25.

Jacob RG, Furman JM, Durrant JD, Turner SM. Panic, agoraphobia,

and vestibular dysfunction. American Journal of Psychiatry 1996;

151:503–12.

Jacob RG, Furman JM, Durrant JD, Turner S. Surface dependence: a

balance control strategy in PD with agoraphobia. Psychosomatic

Medicine 1997;59:323–30.

Jacobson GP, Newman CW. The development of the Dizziness Han-

dicap Inventory. Archives of Otolaryngology, Head and Neck Sur-

gery 1990;116:424–7.

Klein DF. False suffocation alarms, spontaneous panics, and related

conditions. An integrative hypothesis. Archives of General Psy-

chiatry 1993;50:306–17.

Krafczyk S, Schlamp V, Dieterich M, Haberhauer P, Brandt T.

Increased body sway at 3.5–8 Hz in patients with phobic postural

vertigo. Neuroscience Letters 1999;259:149–52.

Leon AC, Shear MK, Portera L, Klerman GL. Assessing impairment

in patients with PD: the Sheehan Disability Scale. Social Psychiatry

and Psychiatric Epidemiology 1992;27:78–82.

Mann CG, Whitney SC, Furman JM. Functional reach and single leg

stance in patients with peripheral vestibular disorders. J Vestibular

Research 1979;5:343–53.

Marks IM, Mathews AM. Brief standard self-rating for phobic

patients. Behavior Research and Therapy 1996;17:263–7.

Miller AD, Yamaguchi T, Siniaia MS, Yates BJ. Ventral respiratory

group bulbospinal inspiratory neurons participate in vestibular-

respiratory reflexes. Journal of Neurophysiology 1995;73:1303–7.

Miller AD, Siniaia MS. Vestibular effects on the upper airway. Brain

Research 1996;736:160–4.

Nattie E. CO2 brainstem chemoreceptors and breathing. Progress in

Neurobiology 1999;59:299–331.

Perna G, Battaglia M, Garberi A, Arancio C, Bertani A, Bellodi L.

Carbon dioxide/oxygen challenge test in PD. Psychiatry Research

1994a;52:159–71.

Perna G, Cocchi S, Bertani A, Arancio C, Bellodi L. Pharmacologic

effect of toloxatone on reactivity to the 35% carbon dioxide chal-

lenge: a single-blind, random, placebo-controlled study. Journal of

Clinical Psychopharmacology 1994b;14:414–8.

Perna G, Bertani A, Arancio C, Ronchi P, Bellodi L. Laboratory

response of patients with panic and obsessive-compulsive disorder

to 35% CO2 challenges. American Journal of Psychiatry 1995;152:

85–9.

Pols H, Zandbergen J, de Loof C, Griez E. Attenuation of carbon

dioxide-induced panic after clonazepam treatment. Acta Psychia-

trica Scandinavica 1991;84:585–6.

Querner V, Krafczyk S, Dieterich M, Brandt T. Patients with soma-

toform phobic postural vertigo: the more difficult the balance task,

the better the balance performance. Neuroscience Letters 2000;285:

21–4.

Sheehan DV, Lecrubier Y, Janavs J, Knapp E, Weiler E, Bonora P.

Mini international neuropsychiatric interview (M.I.N.I.). University

of South Florida Institute for Research in psychiatry, Tampa, Florida

and INSERM- Hopital de la Salpetriere, Paris, France, 1994.

Sklare DA, Stein MB, Pikus AM, Uhde TW. Disequilibrium and

audiovestibular function in PD: symptom profiles and test findings.

American Journal of Otolaryngology 1990;11:338–41.

Spielberger CD, Gorsuch R, Lushere R. STAI manual for the state-

trait anxiety inventory. Palo Alto, CA: Consulting Psychologists

Press, 1970.

Stein MB, Asmundson GJG, Ireland D, Walker JR. PD in patients

attending a clinic for vestibular disorders. American Journal of

Psychiatry 1994;151:1697–700.

Uimonen S, Laitakari K, Kiukaanniemi H, Sorri M. Does posturo-

graphy differenziate malingerers from vertiginous patients? Journal

of Vestibular Research 1995;5:117–24.

Yardley L, Luxon L, Bird J, Lear S, Britton J. Vestibular and pos-

turographic test results in people with symptoms of panic and

agoraphobia. Journal of Audiological Medicine 1994;3:48–65.

Yardley L, Britton J, Lear S, Bird J, Luxon LM. Relationships

between balance system function and agoraphobic avoidance.

Behavior Research and Therapy 1995;33:435–9.

Yardley L, Gresty M, Bronstein A, Beyts J. Changes in heart rate and

respiratory rate in patients with vestibular dysfunction following

head movements which provoke dizziness. Biological Psychology

1998;49:95–108.

Yates BJ, Jakus J, Miller AD. Vestibular effects on respiratory outflow

in the decerebrated cat. Brain Research 1993;629:209–17.

Yates BJ. Autonomic reaction to vestibular damage. Otolaryngology,

Head and Neck Surgery 1998;119:106–12.

Yates BJ, Stocker SD. Integration of somatic and visceral inputs by

the brainstem. Experimental Brain Research 1998;119:269–75.


Panic disorder: the role of the balance system

Documents

Transcript of Panic disorder: the role of the balance system