Research Article

PREVALENCE AND RELATIONSHIP TO DELUSIONS ANDHALLUCINATIONS OF ANXIETY DISORDERS IN

SCHIZOPHRENIA

Philip Tibbo, M.D., F.R.C.P.C.,1n Jennifer Swainson, B.Sc.,2 Pierre Chue, M.R.C. Psych.,3

and Jean-Michel LeMelledo, M.D.4

We investigated the prevalence of anxiety disorders in a sample of individualswith chronic schizophrenia, controlling for anxiety symptoms that may berelated to delusions and hallucinations, and the possible differences in clinicalvariables between the groups. Individuals with a diagnosis of schizophrenia andable to give informed consent were recruited from the community. The MiniInternational Neuropsychiatric Interview (MINI) was administered to bothconfirm the DSM-IV diagnosis of schizophrenia and screen for comorbidanxiety disorders. If a comorbid anxiety disorder was found, its relation to theindividual’s delusions and hallucinations was examined. Clinical rating scalesfor schizophrenia were administered as well as rating scales for specific anxietydisorders where appropriate. Overall, anxiety disorders ranged from 0% [ forPost Traumatic Stress Disorder (PTSD)] to 26.7% [ for generalized anxietydisorder (GAD) and agoraphobia without panic] with lower rates whencontrolled for anxiety symptoms related to delusions and hallucinations. Ininvestigating clinical variables, the cohort was initially divided into schizo-phrenics with no anxiety disorders and those with an anxiety disorder; withfurther analyses including schizophrenics with anxiety disorders related todelusions and hallucinations and those with anxiety disorders not related todelusions and hallucinations. The most consistent difference between all thegroups was on the PANSS-G subscale. No significant differences were found onthe remaining clinical variables. Comorbid anxiety disorders in schizophreniacan be related to the individual’s delusions and hallucinations, though anxietydisorders can occur exclusive of these positive symptoms. Clinicians must beaware that this comorbidity exists in order to optimize an individual’streatment. Depression and Anxiety 17:65–72, 2003. & 2003 Wiley-Liss, Inc.

Key words: anxiety disorders; schizophrenia; comorbidity

DEPRESSION AND ANXIETY 17:65–72 (2003)

1Bebensee Schizophrenia Research Unit, Department of

Psychiatry, University of Alberta Hospital, Edmonton, Alberta,

Canada2Faculty of Medicine and Dentistry, University of Alberta,

Edmonton, Alberta, Canada3Department of Psychiatry, University of Alberta Hospital,

Edmonton, Alberta4Department of Psychiatry, University of Alberta Hospital,

Edmonton, Alberta

nCorrespondence to: Dr. Philip Tibbo, Department of Psychiatry,

University of Alberta Hospital, 1E7.11 WMC, Edmonton, Alberta,

Canada T6G 2B7. E-mail: ptibbo@pop.srv.ualberta.ca

Received for publication 21 December 2001; Accepted 5 October 2002

Published online in Wiley InterScience (www.interscience.wiley.

com). DOI 10.1002/da.10083

&& 2003 WILEY-LISS, INC.

INTRODUCTION

Recently, the scientific community has started toinvestigate psychiatric comorbidity in schizophrenia inmore depth. Evolution of the Diagnostic and StatisticalManual of Mental Disorders (DSM) from the DSM IIIthrough the DSM III-R to the current DSM IV hasallowed for comorbid psychiatric diagnoses to be made.Under the DSM III, diagnosis was dependent on thepremise that one disorder was ‘‘not due to’’ another[American Psychiatric Association, 1980]. For example,a diagnosis of schizophrenia excluded simultaneousdiagnoses of mood, anxiety, and personality disorders[American Psychiatric Association, 1980]. This com-partmentalization of diagnoses was called into questionby Bland et al. [1987] who experimentally disregardedthese DSM III exclusion criteria for schizophrenia andfound that the ‘‘excluded’’ disorders were, in fact,significantly more prevalent among individuals withschizophrenia. With respect to anxiety disorders, Blandet al. [1987], reported that individuals with schizo-phrenia were twenty times more likely to haveobsessive compulsive disorder (OCD), seven timesmore likely to suffer a phobic disorder, and nineteentimes more likely to suffer from panic than normalcontrols. Soon after, revision of the DSM III to theDSM III-R lifted such exclusion criteria and facilitateddiagnoses of other Axis I disorders with schizophrenia.

In more recent years, studies examining comorbidanxiety in schizophrenia have tended to focus primarilyon panic disorders. Prevalence rates ranging from16–46% for panic attacks [Argyle, 1990; Bermanzohnet al., 1996; Cassano et al., 1998; Cutler and Siris,1991; Labbate et al., 1999; Moorey and Soni, 1994;Strakowski et al., 1993] and 11–33% for panic disorder[Bermanzohn et al., 1996; Cassano et al., 1998; Kendleret al., 1996; Labbate et al., 1999; Strakowski et al.,1993] have been reported in the literature.

Argyle [1990] looked beyond panic to also find that5% of his sample showed typical agoraphobic fears,20% had travel fears related to paranoid ideas, andanother 20% displayed social phobia. Similarly, Kend-ler et al. [1996] examined 454 individuals diagnosedwith ‘‘non-affective psychosis’’ (DSM III-R schizo-phrenia, schizophreniform disorder, schizoaffectivedisorder, delusional disorder, and psychosis not other-wise specified) and found prevalence rates of 31% forgeneralized anxiety disorder (GAD), 28% for agor-aphobia, 31% for simple phobia, and 40% for socialphobia. Slightly lower prevalence rates (16% socialphobia, 3% simple phobia, and 0% agoraphobia) werefound by Cassano et al. [1998] who reported on‘‘schizophrenia spectrum’’ patients (DSM IV diagnosisof schizophrenia, schizophreniform disorder, schizoaf-fective disorder, and delusional disorder) using DSMIII-R criteria for anxiety disorders. To our knowledgethere is only one study in the literature that uses DSMIV criteria in individuals with schizophrenia only, toexamine all comorbid anxiety disorders. This group

reported 5% agoraphobia without panic, 17% socialphobia, 12% GAD, and 5% simple phobia in theirsample [Cosoff et al., 1998].

However, there are some difficulties with the studiescompleted to date. Most studies have not (1) surveyedthe prevalence of all the anxiety disorders in the same‘‘true’’ schizophrenic cohort (i.e., excluding schizoaf-fective, non-affective psychosis) using standardizeddiagnostic rating scales for schizophrenia as well asfor anxiety disorders; (2) reported studies of PostTraumatic Stress Disorder (PTSD) prevalence inschizophrenia; (3) dif ferentiated between anxietysymptoms that may be related to delusions andhallucinations and those anxiety symptoms that arenot; or (4) investigated comorbidity and functionaloutcome measures together. We investigated anxietydisorders in individuals with schizophrenia, correctingfor the above deficiencies by (1) investigating themajority of anxiety disorders (including PTSD) in acohort of individuals with DSM-IV schizophrenia, (2)investigating if the anxiety symptoms are related to thedelusions and/or hallucinations of the individualsschizophrenia, and (3) including a measure of func-tional outcome.

PATIENTS AND METHODS

RECRUITMENT

Male and female individuals with a DSM-IVdiagnosis of schizophrenia and able to give informedconsent were recruited by advertisement from anoutpatient population at the University of AlbertaSchizophrenia Clinic as well as a downtown-basedCommunity Living Program (CliP). The individualshad to be considered stable on their dose of anti-psychotics for the 1 month prior to the interview.Inability to give informed consent, significant pasthead injury, significant medical/neurological illness,and active alcohol or drug abuse were consideredexclusion criteria. Research methodology was approvedby the local health research ethics board.

TESTING

Clinical and demographic data were collected oneach individual by interview and corroborated bymembers of his/her treatment team (e.g., case man-agers and treating psychiatrist). This included age,length of schizophrenic illness (defined as time fromfirst hospitalization for schizophrenia to interview),medical and psychiatric history, current/previous med-ications (including dose), and parental socioecomonicstatus. Parental socioeconomic status (PSES) was ratedwith the Hollingshead scale, which ranks individuals ona scale from 1 (families of wealth, education, and top-rank social prestige) to 5 (unskilled and semi-skilledworkers, and elementary education) [Andreasen et al.,1992].

Tibbo et al.66

The Mini International Neuropsychiatric Interview(MINI, Version 4.4) was administered to all individualsto both confirm the DSM-IV diagnosis of schizo-phrenia and screen for comorbid anxiety disorders.Additionally, the Positive and Negative Symptom Scale(PANSS) was administered to all individuals to assessseverity of schizophrenia symptoms and the GlobalAssessment of Functioning (GAF) scale to determineimpact of illness on daily life.

If DSM-IV criteria were met for an anxiety disorderupon administration of the MINI, then its severity wasfurther investigated with the Hamilton Anxiety RatingScale (for GAD), the Liebowitz Social Anxiety Scale(for SP), the Panic and Agoraphobia Scale (for PD),and the CAPS (for PTSD). In addition, if an anxietydisorder was found, the anxiety symptoms were furtherevaluated by clinical interview to determine whetherthey were directly related to the hallucinations anddelusions of the schizophrenic illness, or if it was anunrelated ‘‘pure’’ anxiety. This was accomplished bydetermining the specifics of each individual’s delusionsand hallucinations (e.g., persecutory theme) and then ifany anxiety symptoms co-exist, assessing if they arecaused by and/or temporally associated to the delusionsand hallucinations (e.g., fear of open places due toperceived persecution by CIA agents would beconsidered ‘‘directly related’’).

ANALYSES

As the first part of this study is a descriptive study,descriptive statistics were used. Percentages of indivi-duals with schizophrenia with a comorbid diagnosis ofGAD, SP, PD, or PTSD were calculated. Furtheranalyses included percentages of individuals withschizophrenia with anxiety disorders related to hallu-cinations and delusions (related anxiety) vs. anxietydisorders unrelated to the schizophrenic illness (pureanxiety). Assessment of clinical and demographicdif ferences were achieved using analyses of variance(ANOVA) and two-tailed independent samples t-tests.An alpha of 0.05 was used as the level of significance.

RESULTSSeven women and twenty-five men with schizophre-

nia were interviewed (N¼32 individuals). Of thoseinterviewed, two individuals were later excluded fromthe study after further chart review revealed a history ofepilepsy in one subject and pituitary tumor in another.Of the resulting 30 subjects, there were six women andtwenty-four men.

It was found that GAD was present in 8/30 (26.7%),SP in 7/30 (23.3%), panic disorder with or withoutagoraphobia in 2/30 (6.6%), agoraphobia withoutpanic in 8/30 (26.7%), and 0/30 incidences of PTSD(0%). When those individuals whose anxiety symptomswere related to delusions and/or hallucinations wereexcluded, the rates decreased to the following: GAD in

5/30 (16.7%), SP in 4/30 (13.3%), panic disorder withor without agoraphobia in 1/30 (3.3%), agoraphobiawithout panic in 5/30 (16.7%), and PTSD in 0/30(0%). In individuals with GAD, the average HamiltonAnxiety Score was 14.9, for SP the average Liebowitzscale score was 59.3, and the average Panic andAgoraphobia scale score was 13.7.

The group of 30 patients with schizophrenia wasthen divided into three groups for further analyses: (1)patients with no comorbid anxiety disorder (‘‘noanxiety’’), (2) patients with anxiety disorders relatedto their delusions and hallucinations (‘‘related anxi-ety’’), and (3) patients with anxiety disorders not relatedto delusions and hallucinations (‘‘pure anxiety’’). Thedata were analyzed for dif ferences among these groupson age: the positive syndrome, negative syndrome,general psychopathology, and total scores from thePANSS; the PSES; length of schizophrenic illness; andGAF.

The ‘‘related anxiety’’ and ‘‘pure anxiety’’ groupswere initially combined in order to investigate whetherdifferences existed between patients with no comorbidanxiety disorder and those with anxiety disorders,regardless of whether it was related to the delusionsand hallucinations or not. Patients with comorbidanxiety scored higher on the general pathology scale ofthe PANSS than did patients without anxiety disorders(P¼.024). There was no evidence of further differencesbetween the two groups (Table 1).

ANOVA analysis showed no evidence that the meansof any of the dependent variables differed among thethree groups (Table 2). A Levene’s homogeneity-of-variance test showed that the assumption of equality ofvariances was met in all cases except for the PANSSnegative score (the variances of the ‘‘no anxiety’’ and‘‘related anxiety’’ groups were not equal). However, anindependent samples t-test, with equal variances notassumed, showed that the mean PANSS-N score forthe ‘‘no anxiety’’ group did not dif fer from the mean ofthe ‘‘related anxiety’’ group (t¼0.23, p¼.79).

An independent samples t-test between patients with‘‘related anxiety’’ and those with ‘‘pure anxiety’’ showedno differences on the eight dependent variables.Furthermore, there was no significant dif ferencesbetween these groups on the length of their anxietysymptoms (P¼.94; see Table 3).

Because the sample consisted of 26 men and 4women, and gender dif ferences in prevalence rates forschizophrenia and anxiety disorders exist, an analysesusing gender as a covariate was performed. Noevidence was found for an effect of gender; however,this negative finding would need to be replicated withmore equitable gender sample sizes.

In the cases of patients with ‘‘pure anxiety’’, it wasdetermined that one of these cases began 14 years afterthe onset of schizophrenia, another started about thesame time as the schizophrenia, and the remainder hadonset of anxiety symptoms prior to diagnosisof schizophrenia (average was 9.8 years before

Research Article: Anxiety Disorders in Schizophrenia 67

schizophrenia diagnosis). In the patients with ‘‘relatedanxiety’’, five cases had onset of anxiety symptoms atthe same time as schizophrenia, two cases had onset 8years after schizophrenia diagnosis, with the remainder

having onset of anxiety symptoms, on average, 21.3years prior to their diagnosis of schizophrenia.

No differences were found with respect to anti-psychotic medications among the groups. Medications

TABLE 1. Comparison of descriptive data between individuals with schizophrenia alone and individuals withschizophrenia plus comorbid anxiety disorder (related and not related to delusions and hallucinations)n

Schizophrenia(n¼14)

Schizophrenia withcomorbid (anxietydisorders (n¼16) t P

Age (yr) 38.0 (8.8)a 41.1 (12.7) �0.756 0.46PANSS positive 14.2 (14.8) 13.7 (4.6) 0.306 0.76PANSS negative 17.4 (8.2) 16.3 (5.2) 0.448 0.66PANSS general 26.4 (7.8) 33.3 (7.8) �2.388 0.024PANSS total 58.0 (16.1) 63.8 (15.1) �1.02 0.32PSES 2.9 (0.8) 2.8 (0.9) 0.17 0.87Length of schizophrenic illness (yr) 1.07 (6.5) 12.7 (8.7) �0.684 0.5GAF 61.1 (12.9) 60.2 (8.6) 0.224 0.83

nPANSS, Positive and Negative Symptom Scale; PSES, Parental Socio Economic Status; GAF, Global Assessment of Functioning.aMean (SD).

TABLE 2. Comparison of descriptive data between individuals with schizophrenia alone and individuals withschizophrenia and comorbid anxiety disorders related to delusions and hallucinations, and with comorbid anxietydisorders not related to delusions and hallucinations (‘‘pure anxiety’’)n

Schizophrenia(n¼14)

Comorbid ‘‘relatedanxiety’’ disorders (n¼7)

Comorbid ‘‘pureanxiety’’ disorders (n¼9) F P

Age (yr) 38.0 (8.8)a 45.7 (14.2) 37.4 (10.9) 1.45 0.25PANSS positive 14.2 (4.8) 13.9 (5.3) 13.6 (4.3) 0.053 0.95PANSS negative 17.4 (8.2) 16.6 (4.9) 16.0 (5.7) 0.11 0.89PANSS general 26.4 (7.8) 34.3 (9.8) 32.4 (6.4) 2.87 0.07PANSS total 58.0 (16.1) 66.1 (18.1) 62.0 (13.0) 0.643 0.53PSES 2.9 (0.83) 3.0 (1.0) 2.8 (0.8) 0.142 0.86Length of schizophrenic illness (yr) 10.7 (6.5) 15.1 (11.5) 10.7 (5.7) 0.881 0.43GAF 61.1 (12.9) 59.9 (8.6) 60.4 (9.2) 0.03 0.97

nPANSS, Positive and Negative Symptom Scale; PSES, Parental Socio Economic Status; GAF, Global Assessment of Functioning.aMean (SD).

TABLE 3. Comparison of descriptive data between individuals with schizophrenia and comorbid anxiety disordersrelated and not related to delusions and hallucinationsn

Schizophrenia withcomorbid ‘‘related’’anxiety disorders

(n¼7)

Schizophrenia withcomorbid ‘‘pure’’anxiety disorders

(n¼9) t P

Age (yr) 45.7 (14.2)a 37.4 (10.9) 1.322 0.21PANSS positive 13.9 (5.3) 13.6 (4.3) 0.126 0.9PANSS negative 16.6 (4.9) 16.0 (5.7) 0.211 0.84PANSS general 34.3 (9.8) 32.4 (6.4) 0.455 0.66PANSS total 66.1 (18.1) 62.0 (13.0) 0.533 0.6PSES 3.0 (1.0) 2.8 (0.8) 0.485 0.64Length of schizophrenic illness (yr) 15.1 (11.5) 10.7 (5.7) 1.011 0.33Length of anxiety symptoms 20.0 (18.6) 19.4 (12.0) 0.073 0.94GAF 59.9 (8.6) 60.4 (9.2) �0.131 0.89

nPANSS, Positive and Negative Symptom Scale; PSES, Parental Socio Economic Status; GAF, Global Assessment of Functioning.aMean (SD).

Tibbo et al.68

included risperidone, olanzapine, clozapine, quetia-pine, clopixol, ziprasodone (clinical trial), and fluanxol.

DISCUSSIONThis study confirms that there is a high prevalence of

anxiety disorders in stable outpatient schizophrenics.Our rates of comorbid anxiety disorders are incomparison to that found in the literature to date(Table 4). However, when those individuals whoseanxiety symptoms were found to be related to theirhallucinations and delusions were excluded, theseprevalence rates decreased. The comorbidity rateswere not entirely eliminated though (range,0–17%), highlighting that an individual with schizo-phrenia can suffer from two seemingly entirely separateillnesses.

Our rates are also in comparison to the one otherstudy that used DSM-IV criteria for schizophrenia andscreened for most of the anxiety disorders in thatcohort [Cosoff and Hafner, 1998]. They did notinvestigate PTSD; however in our study, none of oursample met criteria for a diagnosis of PTSD. To ourknowledge, there are no other studies of PTSDprevalence in schizophrenia; however, psychotic symp-toms have been documented in PTSD patients. Ivezicet al. [2000] found that 8/41 PTSD patients hadpsychotic symptoms. These symptoms failed to meetcriteria for schizophrenia and were resistant toneuroleptic treatment. In this case, psychotic symp-toms reflected PTSD-related trauma and appearedrelated to depression [Ivezic et al., 2000]. In a study of53 male combat veterans, David et al. [1999] found that40% suffered from psychotic symptoms, which alsoappeared to be related to depression. Ivezic et al. [2000]found that the hallucinations and delusions of PTSDwere related to the precipitating traumatic event;therefore, it appears that psychotic symptoms may bea feature of PTSD and not of comorbidity. Interest-ingly, Shaner and Eth [1989] took an opposingperspective, suggesting in a theoretical paper thatschizophrenia may be capable of causing PTSD, i.e.,the hallucinations and delusions of schizophrenia mayserve as trauma-inducing events; however our results of0% PTSD in our cohort does not support thishypothesis. Clearly, the relationship between schizo-phrenia and PTSD remains undefined and is an arearequiring further study.

The only dif ferences clinically between the groupsstudied were on the PANSS general psychopathologyscale (those with a comorbid anxiety disorder scoredhigher). This makes intuitive sense as the PANSSgeneral psychopathology section includes anxiety oranxiety-related symptom ratings (e.g., anxiety, tension,and active social avoidance).

Outcome measures reported in schizophrenia andcomorbid anxiety studies are few but include variablesof clinical global impression (CGI) scores, length of

schizophrenic illness, employment, reliance on dis-ability support or sickness benefits, and course ofillness [Cosoff and Hafner, 1998; Cutler and Siris,1991; Emsley et al., 1999; Garvey et al., 1991;Strakowsky et al., 1993]. Most studies support ourfindings that comorbid anxiety does not significantlyaffect outcome. Some studies have even shown morepositive outcome measures in the comorbid groupcompared to individuals with schizophrenia alone[Emsley et al., 1999; Garvey et al., 1991; Strakowskyet al., 1993]. Our sample size in this study did not allowus to examine GAF differences per comorbid diag-nosis, although we recently reported higher GAFscores on individuals comorbid for schizophrenia andobsessive–compulsive disorder compared to individualswith schizophrenia alone [Tibbo et al., 2000]. A futurestudy with a greater sample size, focusing on outcomemeasures per comorbid diagnosis, would enable us todetermine if any of the individual anxiety disordershave more liability on outcome measures in individualswith schizophrenia.

Our sample was obtained by advertisement from twoclinics with similar patient clientele. Although repre-sentative of the general schizophrenia population inEdmonton, one may argue that a sample bias exists inthat those individuals who were more clinically wellwould respond to an advertisement. Although thisshould be consistent in all groups, there may be apossibility that a population of patients with thecomorbid illnesses may be too ill to reply to theadvertisement. Compared to most of the previousstudies, our sample size is quite adequate; however,caution is indicated when the total sample is furtherbroken down for subsequent analyses. An even largerinitial sample size than ours would aid in a morecomplete analyses of the dif ferent variables and/orsubgroups that we were unable to complete forstatistical reasons.

The scientific community needs to now focusattention on the explanation of the association betweenanxiety disorders and schizophrenia. How can thesetwo seemingly distinct entities share such high rates ofcomorbidity? Hoffman [1999] positioned four possibleexplanations for the association: (1) the comorbidityrates are due to methodological artifact, (2) anxietysymptoms cause schizophrenic symptoms, (3) schizo-phrenic symptoms cause anxiety symptoms, or (4)anxiety symptoms and schizophrenic symptoms sharecommon etiologic factors. The methodology of ourpresent study including the use of standardizedassessment of illness using DSM-IV criteria and ratingscales in a stable outpatient population, as well ascontrolling for anxiety symptoms that may be related todelusions or hallucinations controls for possibleexplanations 1 and 3. Thus, we are left with thepossibility that anxiety symptoms may cause schizo-phrenic symptoms or there is overlapping biologicalfactors. While it has been reported that panic attacksmay be associated with increased odds for meeting

Research Article: Anxiety Disorders in Schizophrenia 69

TABLE

4.Comparisonofpreviousschizophrenia/anxietyco

morbiditystudies

Anxietydisorder

comorbidity

Study

Schizophrenia

diagn

osis

Anxiety

measure

Sam

ple

size

Panic

attack

Panic

disorder

Ago

raphobia

Social

phobia

Sim

ple

phobia

Gen

eralized

anxiety

Boyd

etal.,19

84DSM

IIIschizophrenia

DSM

II37

.9%

‘‘panic’’

(disorder

orattacks,unclear)

FF

F

Blandet

al.,19

87DSM

IIIschizophrenia

DSM

III

2030

%‘‘p

anic’’

(disorder

orattacks,unclear)

F63

%‘‘phobia’’in

gen

eral

F

Argyle,

1990

DSM

IIIR

schizophrenia

DSM

IIIR

2035

%F

5%

20%

FF

Cutler

andSiris,19

91Researchdiagn

ostic

criteria

forschizophreniaa

Researchdiagn

osticcriteria

4516

%F

FF

FF

Garveyet

al.,19

91DSM

III

DSM

III

18F

17%

FF

F22%

Strakowskiet

al.,19

93DSM

IIIR

forschizophrenia

spectrum

bDSM

IIIR

10F

20%

FF

FF

Ken

dleret

al.,19

96DSM

IIIR

fornonaffectivepsychosisc

DSM

IIIR

454

F26%

28%

F31%

31%

Bermanzohnet

al.,19

96‘‘chronic

schizophrenics’’

DSM

IV37

16%

11%

FF

FF

Zarate,

1997

DSM

IVforschizophrenia

orschizoaffective

disorder

DSM-IV

60F

19.4%

FF

FF

Cassanoet

al.,19

98DSM

IVforschizophrenia

spectrum

dDSM

IIIR

31F

19%

0%

16%

3%

FCosoff

andHafner,19

98DSM

IVschizophrenia

DSM-IV

60F

5%

5%

17%

5%

12%

Younget

al.,19

99DSM

IVforschizophrenia

DSM-IV

4943

%33

%F

FF

FLabbateet

al.,19

99DSM

IVschizophrenia

DSM-IV

3946

%33

%F

FF

F

a Spitzeret

al.,19

75.

bSchizophrenia

spectrum

included

schizoaffectivedisorder,schizophrenia,andschizophreniform

disorder.

c Nonaffectivepsychosisincludes

schizophrenia,schizophreniform

disorder,schizoaffectivedisorder,delusional

disorder,andpsychosisnototherwisespecified.

dIncludes

schizophrenia,schizophreniform

disorder,schizoaffectivedisorder,delusional

disorder.

Tibbo et al.70

criteria for schizophrenia [Tien and Eaton, 1992] andthat treatment of panic attacks may reduce schizo-phrenic symptoms [Sandberg and Siris, 1988], there isno evidence that these findings are generalizable forthe remaining anxiety disorders. Thus, future studiesshould be directed at investigating a common biologi-cal abnormality or overlapping biological abnormalitiesto explain the rates of anxiety disorders in schizo-phrenia.

One promising area is the suggestion the cholecys-tokinin (CCK) may be the link between anxietydisorders and schizophrenia. The peptide CCK wasoriginally identified as a gastrointestinal hormone butis now known to play an important role as aneuropeptide/neurotransmitter. It is widely distributedthroughout the CNS, with particularly high concen-trations in two major dopamine (DA) systems: themesolimbic and nigrostriatal. Biochemical andelectrophysiological studies indicate a modulatoryrole for CCK on dopamine release, neuronal firing,receptor binding, and post-receptor transductionmechanisms. The precise nature of this modulationis determined largely by the site of action, basalactivity levels, and receptor type. While both CCK-Aand CCK-B receptors have been identified andstudied, the CCK-B receptor is the most predominantin the CNS [for review, see Van Kampen et al.,1997].

CCK has been investigated in both anxiety andschizophrenia separately. Clinically, the administrationof a CCK-B agonist induces panic attacks analogous tospontaneous ones in individuals suffering from panicdisorder and, to a lesser extent, produces anxiety andpanic symptoms in other anxiety disorders. In animalmodels of anxiety, the pretreatment with CCK agonistsand antagonists produced, respectively, anxiogenic andanxiolytic-like action on exploratory paradigms.On the other hand, a significant reduction of CCK,CCK-B mRNA, and isoforms of CCK-B has also beenreported in cortical and limbic structures of individualswith schizophrenia via post-mortem studies [for review,see, e.g., Bourin et al., 1996; Zachrisson et al., 1999].The CCK-B receptor subtype is thought to mediateinhibition of dopamine release and thus a decrease maylead to an increase in DA release, a potential variable inthe development of schizophrenia. Can the role thatCCK plays in anxiety and schizophrenia separately beextended to those individuals who are comorbid withboth illnesses and help explain the significant overlapof symptoms?

In summary, anxiety disorders can be a distinctcomorbid illness in schizophrenia with significantprevalence rates. These anxiety symptoms can also beseparate and not related to the delusions and hallucina-tions that exist in an individual with schizophrenia.Thus, it is important to be able to diagnose andsubsequently treat these symptoms. Future research inthe possible biologic overlap is indicated to further ourunderstanding of this issue.

REFERENCESAmerican Psychiatric Association. 1980. Diagnostic and statistical

manual of mental disorders, 3rd edition. Washington, DC:American Psychiatric Association.

Andreason NC, Flaum M, Arndt S. 1992. The comprehensiveassessment of symptoms and history (CASH): An instrument forassessing psychopathology and diagnosis. Arch Gen Psychiatry49:615–623.

Argyle N. 1990. Panic attacks in chronic schizophrenia. Br JPsychiatry 157:430–433.

Bermanzohn PC, Porto I, Siris SG. 1996. Hierarchical diagnosis inchronic schizophrenia: a clinical study. Proceedings of the 35thannual meeting, American College of Neuropharmacology. SanJuan, Puerto Rico, December 9–13.

Bland RC, Newman SC, Orn H. 1987. Schizophrenia: Lifetimeco-morbidity in a community sample. Acta Psychiatrica Scand75:383–391.

Bourin M, Malinge M, Vasar E, Bradwejn J. 1996. Two faces ofcholecystokinin: anxiety and schizophnenia. Fund Clin Pharmacol10:116–126

Boyd JH, Burke JD Jr, Gruengerg E, Holzer CE 3rd, George LK,Karno M, Stoltzman R., McEvoy L, Nestadt G. 1984. Exclusioncriteria of DSM III: a study of co-occurrence of hierarchy-freesyndromes. Arch Gen Psychiatry 41:983–989.

Cassano GB, Pini S, Saettoni M, Rucci P, Dell’Osso L. 1998.Occurrence and clinical correlates of psychiatric comorbidityin patients with psychotic disorders. J Clin Psychiatry 59:60–68.

Cosoff SJ, Hafner J. 1998. The prevalence of comorbid anxiety inschizophrenia, schizoaffective disorder, and bipolar disorder.Austral N Z J Psychiatry. 32:67–72.

Cutler JL, Siris SG. 1991. Panic-like symptomatology inschizophrenic and schizoaffective patients with postpsychoticdepression: observations and implications. Comp Psychiatry32:465–473.

David D, Kutcher GS, Jackson EI, Mellman TA. 1999. Psychoticsymptoms in combat-related posttraumatic stress disorder. J ClinPsychiatry 60:29–32.

Emsley RA, Oosthuizen PP, Joubert AF, Roberts MC, Stein DJ.1999. Depressive and anxiety symptoms in patients with schizo-phrenia and schizophreniform disorder. J Clin Psychiatry 60:747–750.

Garvey M, Noyes R Jr, Anderson D, Cook B. 1991. Examination ofcomorbid anxiety in psychiatric inpatients. Comp Psychiatry32:277–282.

Hofmann SG. 1999. Relationship between panic and schizophrenia.Depression Anxiety 9:101–106.

Ivezic S, Bagaric A, Oruc L, Mimica N, Ljubin T. 2000. Psychoticsymptoms and comorbid psychiatric disorders in Croatian com-bat-related posttraumatic stress disorder patients. Croat Med J41:179–183.

Kendler KS, Gallagher TJ, Abelson JM, and Kessler RC. 1996.Lifetime prevalence, demographic risk factors, and diagnosticvalidity of nonaffective psychosis as assessed in a US communitysample: the national comorbidity survey. Arch Gen Psychiatry53:1022–1031.

Labbate LA, Young C, Arana GW. 1999. Panic Disorder inSchizophrenia. Can J Psychiatry 44:488–490.

Moorey H, Soni HD. 1994. Anxiety symptoms in stable chronicschizophnrenics. J Ment Health 3:257–262.

Sandberg l, Siris SG. 1987. ‘‘Panic disorder’’ in schizophrenia: a casereport. J Nerv Ment Dis 175:627–628.

Shaner A and Eth S. 1989. Can schizophrenia cause posttraumaticstress disorder? Am J Psychotherapy 43:588–597.

Research Article: Anxiety Disorders in Schizophrenia 71

Strakowski SM, Tohen M, Stoll AL, Faedda GL, Mayer PV,Kolbrener ML, Goodwin DC. 1993. Comorbidity in psychosis atfirst hospitalization. Am J Psychiatry 150:752–757.

Tibbo P, Kroetsch M, Chue P, Warneke L. 2000. Obsessive-compulsive disorder in schizophrenia. J Psychiatric Res 34:139–146.

Tien AY, Eaton WW. 1992. Psychopathologic precursors andsociademographic risk factors for the schizophrenia syndrome.Arch Gen Psychiatry 49:37–46

Van Kampen J, Stoessl AJ. 1998. The effects of CCK-4 on dopamineD1 agonist-induced grooming are blocked by a CCK-A receptor

antagonist: evidence for a novel CCK receptor subtype. Neuro-pharmacology 36:1679–1688.

Zachrisson O, de Belleroche J, Wendt KR, Hirsch S, Lindefors N.1999. Colecystokinin CCK(B) receptor mRNA isoforms:expression in schizophrenic brains. Neuroreport 10:3265–3268.

Zarate R. 1997. The comorbidity between schizophrenia and anxietydisorders. Proceedings of the 31st annual meeting of theAssociation for the Advancement of Behavior Therapy, MiamiBeach, FL, November 13–16.

Tibbo et al.72