Multicultural studies and the nature of schizophrenia: a review

162 Journal of the Royal Society of Medicine Volume 80 March 1987

Multicultural studies and the nature of schizophrenia: a review

Assen Jablensky MD DMSc Division ofMental Health, World Health Organization, Geneva, Switzerland

Keywords: schizophrenia aetiology, cross-cultural aspects; incidence; course and outcome

Schizophrenia, a condition of obscure origins and no

established aetiology, pathogenesis and pathology,continues to be a major focus of psychiatric researchand public health concern. The diagnosis of schizo-phrenia accounts today for well over 50% of the'long-stay' psychiatric hospital populations in manyindustrialized countries, and extrapolations from thecurrent incidence rates and demographic trends leadto predictions offurther increases ofthe total numberof cases by the end of the century, especially in ThirdWorld countries'.

The uncertain nosological status ofschizophreniaThere are few conditions in medicine that have beeninvestigated with such persistence and with so fewtangible returns over decades as has schizophrenia.Notwithstanding the current availability of pharma-cological means for control of some of its floridsymptoms, and in spite of the impressive advancesof the modern technology for biological research,no disease marker and no laboratory test are yetavailable for the identification of schizophrenia, thediagnosis relying entirely on clinical judgment andconvention. In practice this means that the identifi-cation of the disorder depends on the psychiatrist'sinterpretation of the patient's reported subjectiveexperience, on observations of his behaviour andform of communicating, and on descriptions of thepattern in which the symptoms evolved.The disease entity theory which underlies current

diagnostic and classificatory approaches to schizo-phrenia is entirely inferential, drawing upon circum-

stantial rather than direct evidence. The argumentsadvanced today in support of the disease theory ofschizophrenia fall roughly into four groups: (1)genetics (ifa family history is present, the degree ofanindividual's genetic kinship with a person diagnosedas schizophrenic is a reliable predictor of that indi-vidual's risk of developing the condition himself); (2)courseandoutcome (insome cases schizophrenia tendsto result in some degree of personality alterationwhich is so characteristic that it can serve as a cri-terion for catamnestic verification of the diagnosis);(3) treatment response (psychopharmacologicalagents, especially those acting on the brain'sdopaminergic systems, can suppress many of thesymptoms ofschizophrenia more effectively than anyother intervention); (4) cerebralpathology (structuralbrain abnormalities can be found in a certain propor-tion of schizophrenic patients investigated with thenewer brain imaging techniques).None of the above features is either necessary or

sufficient to establish a disease status for schizo-phrenia. The genetic findings are compatible withdifferent interpretations. Moreover, no familyhistoryof the disorder can be ascertained in the majority of

cases of schizophrenia, and even when a genetic'load' is undoubtedly present (as in the monozygoticco-twin of a schizophrenic or in the offspring of twoschizophrenic parents) the risk ofdeveloping the dis-order falls short of50%. The treatment response is nobetter evidence that we are dealing with a diseaseentity, and the course and outcome criterion is weakas it is not evident why the same disease should notproduce different outcomes, or why the same patternof course should not occur in different diseases. Thesignificance of the structural brain abnormalities isequally open to conflicting interpretations.Leaving aside as extremely implausible the

attempts to formulate a 'non-disease' theory ofschizo-phrenia as a kind of social deviance or a sequel ofsocial labelling, there are at least two alternativesto the single disease model of schizophrenia. Thediathesis-stress model regards schizophrenia as acontinuum of abnormal neurophysiological andbehavioural responses to environmental stress inindividuals possessing some kind of constitutionalvulnerability. The other alternative is the view thatschizophrenia is not a single disease but a collectionof different syndromes or diseases, each having acharacteristic clinical pattern, prognosis and cause.Although the alternative conceptualizations ofthe

nature of schizophrenia have very different impli-cations for the type of research that might advancethe understanding ofthe causes ofthe disorder, thereis at present no decisive test to discard any of thecompeting theories. The essential symptoms ofschizo-phrenia appear to be the expression of disturbancesat the highest and least well understood levels ofintegration ofspecificallyhumanneuropsychologicalfunctions - the delimitation of the self from the outerworld, the encoding and decoding of informationcritical for social communication, and other high-order information processing for which there is noanimal model to assist experimental research. Beinga disorder with a relatively low incidence, onset inyoung adult age, and quite unpredictable evolutionof the clinical manifestations over years or decades,schizophrenia is refractory even to the most powerfulresearch strategies, such as prospective cohortinvestigations. Much of the costly neurobiologicalresearch is also bound to remain unproductive,unless planned and conducted in the context of abroader strategy based on a nosological construct.

Cross-cultural and epidemiological approachesCan cross-cultural and epidemiological studies - or across-breed of the two approaches - provide usefulguidance and testable hypotheses for clinical andlaboratory research into the nature ofschizophrenia?Both the epidemiological and cross-cultural

approaches to the study of schizophrenia have theirbirthdates around the turn of the present century.

Paper read toSection ofPsychiatry,12 November 1985

0141-0768/87/030162-06/$02.00/0i 1987The RoyalSociety ofMedicine

Journal of the Royal Society of Medicine Volume 80 March 1987 163

The origins of both were connected in one way oranother with the name of Emil Kraepelin. At theinstigation ofKraepelin, in 1895 Jenny Koller carriedout the first community survey ofpsychotic disordersin two random samples ofsubjects in the Swiss cantonof Zurich2. In 1904, after a trip to Java, Kraepelinpublished his own cross-cultural observations ondementia praecox and manic-depressive illness3and advocated 'comparative psychiatry' as a newapproach to the understanding of the causes andmechanisms of mental disorders. He warned, how-ever, that 'reliable comparison is, of course, onlypossible if we are able to draw clear distinctionsbetween identifiable illnesses, as well as betweenclinical states; moreover, our clinical conceptsvary so widely that for the foreseeable future suchcomparison is possible only if the observations aremade by one and the same observer'.Such variation of concepts and points of view

could hardly be avoided at the time, and cross-cultural studies and epidemiological research tookdifferent and separate roads. Cross-cultural researchand ethnopsychiatry became associated with apsychoanalytically oriented variety of culturalanthropology which principally sought evidence thatpsychodynamic mechanisms were cultural univer-sals. Although psychosis was not the central concernof cultural anthropology, the conceptualizations ofschizophrenia produced in its framework were some-times original and striking. Thus to Devereux4schizophrenia was 'the typical ethnic psychosis ofcomplex civilised societies', and Batesons proposedthe 'double-bind' hypothesis according to which aculturally determined communication conflict was atthe root of schizophrenic disturbances. One of thecorollaries of the ethnopsychiatric view was theexpectation that the disorder would not be univer-sally distributed and that different cultures wouldgenerate different quantities of schizophrenia.

In contrast, the epidemiological approach pro-ceeded from amore conventional clinical definition ofthe disorder and attempted to measure its frequencyin different societies in terms of prevalence, morbidrisk and, less often, incidence. Most prevalencestudies, whether based on total community surveys,population samples, or hospitalized cases, have pro-duced similar rates ranging, with a few exceptions,from 2 to 10 per 1000 in different countries6.Few studies have attempted to determine the rates

of manifestation of schizophrenia in sufficientlycontrasting cultural groups. To mention only a fewexamples, Rin and Lin7 found that although the life-time incidence of all mental disorders among theaborigines ofTaiwan (China) was the same as amongthe Chinese population of the island, the rate ofschizophrenia in the former was significantly lower.In a study of first admissions for schizophrenia inMauritius, Murphy and Raman' observed somemarked differences between the three major ethnicgroups on the island- the Indian Moslems, the HinduIndians and the non-Indians - the first having thelowest and the last the highest first admission rates.In the western regions of Yugoslavia, Crocetti et al.9identified 2 geographical areas where the prevalenceof schizophrenia and other psychoses was nearly 3times as high as in the rest of the country.The mapping of the frequency of schizophrenic

disorders against cultural, ethnic and geographicalvariation has been slow and of limited success

because of two methodological obstacles. First,comparisons ofresults of individual studies could notbe made with confidence because of the lack of com-parability of case definitions, case-finding methodsand assessment techniques. Secondly, most studiesfocused principally on identifying and enumeratingcases and, as a rule, did not provide adequate clinicaldata on symptomatology and course in a way thatallowed secondary analysis. In the absence of suchinformation, it is difficult to judge how similar ordifferent were the patient populations covered bysuch epidemiological studies.

Schizophrenia research in the programme oftheWorld Health OrganizationThe programme of collaborative clinical and epi-demiological research into schizophrenia andrelated disorders, undertaken by the World HealthOrganization in the late 1960s and continuing intothe present, aimed to overcome some of the method-ological obstacles and clear the way for more reliablecomparative investigations in different populations.The progamme involved the participation of psychia-trists and other investigators in 20 centres in 17countries. The research strategy ofthe programme ischaracterized by the following general features:

(a) The use of standardized instruments for case-finding,history-taking and mental state assessment, in equivalenttranslations into the local languages of the populations ofthe study areas.(b) Data collection by highly qualified psychiatrists; exten-sive training of the investigators in the reliable use of theinstruments and monitoring of the rliability of patientassessment through joint reliability exercises and exchangeof tapes or transcripts of clinical interviews.(c) A two-tier diagnostic classification of the clinical infor-mation, involving a clinical diagnosis by the local teamof investigators and assignment of the cases to referencediagnostic classes by a computer programme (CATEGO10)at the study headquarters in Geneva.(d) Multiple assessments of the patients, with follow-upexaminations at intervals of one, two, and five years.

The programme has included three studies (Table1). The first (1969-1977) was the International PilotStudy of Schizophrenia'112, involving nine centresin Africa, Asia, Europe and North America and atotal of 1202 patients aged 15-44, selected on the basisof screening criteria which required the presence ofpsychotic symptoms and excluded gross cerebralpathology, chronic psychotic conditions of longduration, sensory defects and mental retardation.The majority of the patients (811) had a clinicaldiagnosis of schizophrenia and the remaining 391,selected for comparison purposes, had diagnoses ofaffective or reactive psychoses, neuroses and person-ality disorders. Each patient had a detailed standard-ized clinical examination at the point of inclusionand a complete standardized reassessment two yearsand five years later (82% of the initial cohort had atwo-year follow up and 75% had a five-year follow up).The second WHO study was designed to explore

more closely the manifestations and course ofbehavioural impairments and social disability inschizophrenic patients of recent onset. It included520 patients in seven countries who were examinedinitially and also at one-year and two-year follow-upassessments (one of the centres carried out periodicre-examinations every six months, and two centresundertook further re-assessment at five years). In


Table 1. The WHOprogramme ofcross-cultural research in schizophrenia (1967-85)

International Pilot Study of WHO collaborative study on Determinants ofoutcome ofSchizophrenia (IPSS) psychiatric disability severe mental disorders

Number of centres 9 7 12Countries China (Taipei), Colombia, Bulgaria, Federal Republic Colombia, Czechoslovakia,

Czechoslovakia, Denmark, ofGermany, Netherlands, Denmark, India, Ireland, Japan,India, Nigeria, UK, USA, Sudan, Switzerland, Turkey, Nigeria, UK, USA, USSRUSSR Yugoslavia

Number of patients 1202 520 1379Main areas assessed Mental state (PSE), past Mental state (PSE), past Mental state (PSE), past history,

history, social description, history, sociodemographic course and outcome, disabilitycourse and outcome description, disability in in social roles, stressful life

social roles, behavioural events, expressed emotion,impairments, pattern of perception of illness, familycourse functioning

Diagnosis Clinical (ICD-8), computer Clinical (ICD-9), computer Clinical (ICD-9), computer(CATEGO), statistical (CATEGO) (CATEGO), DSM-III (someclusters centres)

Follow up 2 years, 5 years 1 year, 2 years, 5 years 1 year, 2 years

addition to the Present State Examination (PSE)'0,which was used as a standard mental state interviewthroughout the WHO programme, two new instru-ments were developed especially for this study. Oneof them was the Psychological Impairment RatingScales (PIRS), designed to describe communicationand social skills deficits in everyday behaviour (manyof these disturbances could also be seen as 'negative'symptoms of schizophrenia). The other instrument,the WHO Disability Assessment Schedule (DAS),served to elicit and rate information on social roleperformance and the environmental factors thatmight influence it"3.The third and most recent study (1978-1984),

Determinants of Outcome of Severe Mental Dis-orders, had a more complex design and included atotal of 1379 patients assessed at 12 research centresin 10 countries. The core of the project was an epi-demiological case-finding and clinical study whichaimed to provide data on the comparative incidenceof schizophrenic disorders in different geographicalareas, the frequency ofvarious patterns ofcourse, andthe relationship between epidemiological measuresand alternative diagnostic classifications of thecases"4. It involved a continuous 2-year surveillanceof psychiatric and other medical services, socialagencies, and alternative sources of care such astraditional or religious healers, in defined catchmentareas. The case-finding method involved an activesearch for individuals making first lifetime contactswith such facilities because of manifest or possiblepsychotic illness. A screening procedure was appliedto identify subjects with probable schizophrenic orother non-affective illness, and it was followed by afull mental state, history and diagnostic assessmentof the eligible cases. Follow-up examinations werecarried out one year and two years after the initialscreening. The cohorts collected in this manner wereconsidered to approximate a representative incidencesample. The great majority ofthe patients (86%) wereidentified within the first 12 months ofthe onset ofthedisorder and practically all patients were included inthe study within 3 months of their first contact withany service - i.e. before treatment effects or socialfactors could pathoplastically alter the manifes-tations and course of the illness. In addition to the

standard clinical, social and diagnostic assessment,subgroups of the patients participated in specialinvestigations, such as an exploration of life eventspreceding the onset of symptoms'5, a study of theeffect of expressed emotion in a relative on theprobability of relapse"6, a study of the perception ofthe patient's behaviour by the social environment,and an investigation-of the rate of development ofvarious social role dysfunctions.

Results ofthe WHO studiesThe main findings of the WHO multi-centre pro-gramme can be summarized as follows:

(1) The major psychopathological syndromesdefining the clinical entity of schizophrenia since itsdelimitation by Kraepelin'7 and Bleuler'8 have beenfound tooccurin all thepopulations and geographicalareas covered by the WHO investigations. Althoughno single symptom was invariably present in everypatient and in every setting, the clinical picturesassociated with a diagnosis of schizophrenia wereremarkably similar at the level of symptom profiles(Table 2). Patients diagnosed as schizophrenic tendedto have high scores on lack ofinsight, suspiciousness,delusional mood, delusions or ideas of reference andpersecution, flatness of affect, auditory halluci-nations, and the delusion of being controlled by anexternal agency.The application of CATEGO demonstrated a high

degree ofagreementbetweenthe centrediagnosis andthe computer classification of the cases (an averageof87% in the IPSS and between 63% and 95% of thecases in the different centres in the Outcome Study).Varying proportions (between 31% and 85%, anaverage of 56%) of the patients meeting the generalcriteria of a non-affective fimctional psychosisand diagnosed by clinicians as schizophrenic alsoexhibited one or more of the 'first-rank' symptomsconsidered by Schneider'9 as reliably distinguishingschizophrenia from other non-organic psychotic ill-nesses. These symptoms appeared to define a sub-population of schizophrenic patients characterizedby a generally high frequency and intensity of'positive' psychotic symptoms which manifestedgreat similarity across the cultures.

(2) Against a background of overall similarity in


Table 2. International Pilot Study ofSchizophrenia (IPSS).Ten most frequently positive 'units of analysis' in patientswith diagnosis of paranoid schizophrenia in Aarhus,Agra, Cali, Ibadan, London, Moscow, Prague, Taipei andWashington

1) Lack of insight2) Suspiciousness3) Delusions of

persecution4) Delusions of

reference5) Ideas of reference6) Uncooperativeness7) Inadequate

description8) Delusional mood9) Flatness of affect

10) Auditoryhallucinations

55% (Was) - 100% (Agr, Mos)67% (Pra) - 93% (Agr)

60% (Cal) - 93% (Agr)

54% (Mos)46% (Aar)28% (Lon)

32% (Lon)36% (Pra)41% (Iba)

- 73% (Agr)- 86% (Tai)- 82% (Aar)

- 83% (Was)- 75% (Cal)- 68% (Lon)

31% (Was) - 64% (Pra)

the presentation of schizophrenic illnesses in differ-ent cultures, there were differences among the indi-vidual centres and, in particular, between the studyareas in developing countries and in industrializedcountries. Such differences are of interest becausethey may reflect pathoplastic effects ofculture ratherthan chance variation. First, a significantly greaterproportion of patients in the developing countrieswho met the symptomatological criteria for schizo-phrenia had an acute onset ofthe disorder. Secondly,patients in developing countries exhibited feweraffective symptoms (e.g. depression) as part of theinitial manifestations ofschizophrenic illnesses thantheir counterparts in European, North American andJapanese centres. Thirdly, patients in developingcountries had higher scores on auditory and visualhallucinations. However, these features ofthe presen-tation of psychotic illnesses in developing countriesdid not cluster together as a separate syndrome,different from the 'central' schizophrenic syndrome.On the contrary, they were present in associationwith the 'nuclear' syndrome defined by Schneiderianfirst-rank symptoms.

(3) One of the striking findings of the follow-upphase ofthe IPSS was the contrast between the initialsymptomatological similarity of the schizophrenicpatients both within and across centres, and thegreat variety of forms of course and outcome whichtheir illnesses took over the subsequent five years.Generally, there were three characteristic types ofcourse. A proportion of the patients had a singleacute or subacute psychotic episode which wasfollowed by a complete recovery without furtherattacks. At the other extreme were the patients whoremained ill throughout the follow-up period andhad chronic psychotic symptoms and severe socialincapacitation. An intermediate group comprisedpatients who had several psychotic attacks withinterposed remissions which could be complete orpartial. In the IPSS study population, 27% of theschizophrenic patients fell into the first group, 26%into the second, and 47% into the third. However,the frequency of the three pattems of course variedsignificantly among the centres. While 58% of theNigerian patients and 51% ofthe Indian patients hada single psychotic episode followed by a completeremission, the corresponding percentages for thistype of course in the remaining centres range

o I i I 1

Good 1 2 3 4 5 Pooroutcome Categories of outcome outcome

Figure 1. Distribution of 233 followed-up schizophrenicpatients in developing countries and 295 followed-up schizo-phrenic patients in developed countries over 5 categories of2-year overall outcome. (Reproduced from Sartorius et al.20,with kindpermission)

between 6% in Denmark and 27% in China (Taipei).In contrast, 50% of the patients in Denmark, 47% inthe USA, and 30% in theUK and Czeckoslovakia hada chronic unremitting psychotic illness, comparedwith only 7% in Nigeria and 20% in India. An index of'overall outcome', combining the score on pattern ofcourse, the summary duration of psychotic episodes,and the scores on quality of remissions and degreeof social impairment, demonstrated that the IPSSpatients in the developing countries had a signifi-cantly better outcome than the patients in thedeveloped countries (Figure 1).A number of predictors associated with each of the

three patterns of course were identified by multi-variate statistical analysis. The remitting type waspredicted by an acute onset, absence of previousepisodes of psychiatric illness, and stable familybackground. The chronic pattern representedusually the extension of an illness which had startedinsidiously in the past, in a socially withdrawnperson with a long history of abnormal behaviourwho was also likely to be single, divorced, separatedfrom spouse, or widowed. The periodic recurrentpattern was more frequent in females and the initialepisodes of the illness were characterized by anadmixture of schizophrenic, affective and neuroticsymptoms.While none ofthese predictors would be unexpected

to clinicians, the three different patterns of coursecorrelated poorly with the initial diagnostic classifi-cation of the cases. Neither the ICD subtypes ofschizophrenia, nor the CATEGO classification of'nuclear' and non-'nuclear' schizophrenic syndromespredicted adequately the subsequent pattern ofcourse, with the exception of the distinction betweena subgroup combining simple and hebephrenicschizophrenia and the subgroup ofcases diagnosed asschizoaffective.This finding underscores the difficulty of reconcil-

ing the two main approaches to the diagnosis ofschizophrenia - one emphasizing primarily courseand outcome, and the other giving priority to thecross-section of psychopathological manifestations.


Although the existence, within the IPSS population,of a subgroup of patients with acute transient psy-

chotic illnesses which bear no intrinsic relationshipto schizophrenia but mimic some of its symptoms,cannot be excluded, the data support the notionof a pathoplastic effect of the social and culturalenvironment onthecourse ofschizophrenic disorders.

(4) The incidence rates ofschizophrenia, measuredin terms of annual rate of first lifetime contacts withany type of service, are comparable in culturallydistant populations and vary by only a factor ofthreebetween areas with high rates (e.g. a rural area near

Chandigarh, India) and areas with low rates (thecounty of Aarhus, Denmark). Although the differ-ences observed among the six centres in the OutcomeStudy which achieved a fairly complete coverage ofall first contacts were statistically significant(P< 0.05), their magnitude was not of an orderthat would suggest major cultural contrasts in theincidence of schizophrenia. Even more importantly,the application to the data of a restrictive diagnosticcriterion (the CATEGO class S+ defining a 'central'schizophrenic subgroup) resulted not only in lowerincidence rates as expected but also in a markedreduction of the inter-area variation of the rates(Figure 2). This was contrary to the prediction that,as the mean rate of occurrence of a disease declines,the spread of the individual rates for the differentcentres would increase (because the chance additionor omission of a single case would result in a greaterpercentage difference in the smaller sample). It couldbe inferred, therefore, that the 'nuclear' schizo-phrenic syndrome may indeed be occurring at an

almost uniform rate in different populations.This inference was further supported by the non-

random distribution of the age at onset and themarked differences in this respect between males andfemales (in the absence ofsex differences in the over-

all incidence rates when the entire age range 15-54

500

'Broad" definition:400 Clinical diagnosis or

Catego S, P, 0

"~~1'1R octrir_tivP- d gafinitinn --

Co sco ooi <o s oCD o co

c oZ C zo c z z=0 0<0 00m<

Figure 2. Incidence rates per millon population age 15-54(both sexes):for the 'broad'andfor the 'restrictive definition ofschizophrenia. (Reproduced from Sartorius et al.14, withkind permission)

was considered). In virtually all centres, the illnessonsets in males showed a peak in the age groupsyounger than 24, while in females the onsets tendedto cluster in older age groups. With the application ofthe stricter CATEGO S+ criterion these distinctionsbecame even sharper.

Implications ofthe WHO data for conjecturesabout the nature of schizophreniaThe selected findings of the WHO research pro-gramme presented here are still of a preliminarynature. Data analyses now in progress may addimportant new material, although it is unlikely thatthe essential configuration of the results will bechanged. With this assumption, itmay be tempting toreflect on the possible implication of the findings forthe general disease theory of schizophrenia.A first corollary of the findings is that any extreme

position of cultural relativism as regards the identi-fication of schizophrenia in different populationswould be untenable. The fact that a specific pattern ofsymptoms can be identified reliably in widely varyingcultural settings, and that this pattern correspondsfairly well to a generally accepted clinical descrip-tion ofthe schizophrenic syndrome, is only one ofthefindings that has to be accounted by the relativistposition. Another is the clear-cut symptomatologicaldistinction, observed in all settings, between con-ditions diagnosed as schizophrenic and those ofothertypes, e.g. manic-depressive illness. A third obser-vation is the similarity of certain characteristicmanifestations of schizophrenia across differentcultures. Considering the variety of social norms,beliefs, attitudes and techniques for coping withstress which exist in different cultures, the similarityof the subjective experience of certain schizophrenicsymptoms is quite striking. For example, some ofthe'first-rank' symptoms refer to highly specific intra-psychic phenomena which could serve as a 'marker'for the identification of the 'nuclear' schizophrenicsyndromes. It is difficult to think of a common'cultural' reason for acutely disturbed subjects indifferent cultures to experience hallucinatory voicesdiscussing them precisely in the third person orcommenting on their every action and thought; feeltheir thoughts being stopped, taken away, 'read' bysome alien agency, or 'broadcast' at large. Yet thisis what patients in cultures as different from oneanother as Denmark and Nigeria describe, usingalmost identical words or phrases. Unless this obser-vation is shown to be an artefact of the interviewingtechnique - which is unlikely - it suggests stronglythat the forms of schizophrenic experience, i.e. thespecific disorders of perception, thought, self-imageand ideation, have a common pathophysiologicalbasis and are universal.A second corollary is that the finding of similar

incidence rates of the schizophrenic syndrome indifferent populations need not necessarily imply thatthe causes of schizophrenia are exclusively 'biologi-cal'. Considering the marked genetic, constitutional,nutritional, and other biological differences amongthe populations studied, the emergence of similarrates of any complex disorder would be expected.Insofar as a methodological bias in case-findingand ascertainment can be excluded, schizophreniaappears to behave differently from other multi-factorial diseases like diabetes or ischaemic heartdisease, whichshowenormous variations inincidence


in the same populations. If schizophrenia is not asingle disease of uniform aetiology and pathology,but rather a broadly defined syndrome that may ariseas a response to a variety of pathological processesor developmental anomalies - some of them withstrong genetic contribution and some the result ofenvironmental factors - then the comparable rates ofmanifestation could be seen as the expression of asimilarly distributed liability for a 'schizophrenic'type of response rather than as a reflection of asimilar distribution of identical primary causes. Apossible analogy may be found in epilepsy, whichseems to be one of the few other disorders occurringwith similar rates in very different populations,the ictal discharge being a response modality of acertain neurophysiological organization which canbe activated by a variety of lesions and stimuli.A third corollary is that research into the con-

ditions wh.ich may facilitate or inhibit the mani-festation of a schizophrenic syndrome or 'reactiontype' may be a fruitful approach. It is well known thatschizophrenic symptoms do arise in association withcentral nervous system intoxications (e.g. LSD, PCP,amphetamine), temporal lobe epilepsy, and a varietyof cerebral degenerative diseases, but little is knownabout the determinants of individual susceptibilityto schizophreniform illness in some cases, to acuteconfusional episodes in other cases, and to cognitivedisorders of a dementing type in yet another propor-tion of subjects. Even less is known about thepsychosocial and cultural conditions that may eitherpreclude or augment the probability of a schizo-phrenic response to pathological lesions. The WHOstudies have only pointed to a possibility that intechnologically less complex cultures the chronicdeteriorating forms of schizophrenia may be less fre-quent than in societies imposing upon their memberscomplex, conflicting and potentially disorientingcognitive requirements. Knowledge about the occur-rence ofschizophrenia in radically different societies- e.g. in pre-literate cultures or hunter-gatherergroups - is lacking and extremely difficult to obtain.The possibility, however, cannot be excluded that theapplication of the modern research technology insuch explorations may bring us closer to a newvantage point from which to investigate the elusivenature of schizophrenia.This may involve the questioning of certain basic

assumptions, an example of which was provided byKraepelin who, towards the end of his career21, putthe problem of schizophrenia into a perspectivewhich was fundamentally different from his ownearlier view. In his article 'Die Erscheinungsformendes Irreseins' he proposed that:

'the affective and schizophrenic forms ofmental disorder donot represent the expression of particular pathological pro-cesses, but rather indicate the areas of our personality inwhich these processes unfold ... It must remain an openquestion whether this isdue to generalhuman psychologicalmechanisms operating in combination with pathologicalchanges, or whether hereditary factors make certain areasmore susceptible and accessible to pathological stimuli .. .The various syndromes of illness may be compared with thedifferent registers of an organ, any ofwhich may be broughtinto play according to the severity and extent of the patho-logical changes involved. They impart a characteristic toneto the illness quite irrespective of the mechanism which hasbrought them into play . .. Schizophrenic symptoms are byno means limited to dementia praecox. We find them also in

varying degree in many morbid processes in which there iswidespread destruction ofnerve tissue ... There is no doubt,however, that schizophrenic symptoms may also occurwithout any damage to cerebral tissue'.

Sixty-five years later, we find ourselves challengedto define a better research agenda for biological andepidemiological psychiatry.

References1 Kramer M. The rising pandemic of mental disorders

and associated chronic diseases and disabilities. ActaPsychiatr Scand 1980;62(Suppl. 285):382-97

2 Koller J. Beitrag zur Erblichkeitsstatistik der Geis-teskranken im Canton Zurich. Archiv fur Psychiatrieund Nervenkrankheiten 1895;27:268-94

3 Kraepelin E. Comparative psychiatry (1904). In:Hirsch SR, Shepherd M, eds. Themes and variations inEuropean psychiatry. Bristol: Wright, 1974;

4 Devereux G. Essais d'ethnopsychiatrie gne'nrale. Paris:Gallimard, 1970

5 Bateson G, Jackson DD, Haley J, Weakland JH.Towards a theory of schizophrenia. In: Bateson G, ed.Steps to an ecology ofmind. St Albans: Paladin, 1973:

6 Jablensky A, Sartorius N. Culture and schizophrenia.Psychol Med 1975;5:113-46

7 Rin H, Lin TY. Mental illness among Formosanaborigines as compared with the Chinese in Taiwan.Journal ofMental Science 1962;108:134-46

8 Murphy HBM, Raman AC. The chronicity of schizo-phrenia in indigenous tropical peoples. Results in atwelve-year follow-up survey in Mauritius. Br J Psy-chiatry 1971;118:489-97

9 Crocetti GM, Lemkau PV, Kulcar Z, Kesic B. Selectedaspects of the epidemiology of psychoses in Croatia,Yugoslavia. Am JEpidemiol 1971;94:126-34

10 Wing JK, Cooper JE, Sartorius N. The measurementand classification of psychiatric symptoms. London:Cambridge University Press, 1974

11 World Health Organization. Report ofthe InternationalPilot Study ofSchizophrenia, Vol 1. Geneva: WHO, 1973

12 World Health Organization. Schizophrenia. An Inter-national Follow-Up Study. Chichester: Wiley, 1979

13 Jablensky A, Schwarz R, Tomov T. WHO CollaborativeStudy on Impairments and Disabilities Associated withSchizophrenic Disorders, Acta Psychiatr Scand 1980;62(Suppl. 285):152-63

14 Sartorius N, Jablensky A, Korten A, Ernberg G, CooperJE, Day R. Early manifestations and first-contact inci-dence of schizophrenia in different cultures. PsycholMed 1986 (in press)

15 Day R, Nielson JA, Korten A, et al. Stressful life eventspreceding the acute onset of schizophrenia: a cross-national study from the World Health Organization.Cult Med Psychiatry 1986 (in press)

16 Wig NN, Menon DK, Bedi H, et al. The cross-culturaltransfer of ratings of relatives' expressed emotion. Br JPsychiatry 1986 (in press)

17 Kraepelin E. Psychiatrie. Ein Lehrbuch fur Studirendeund Aerzte. 5. Aufl. Leipzig: Barth, 1896

18 Bleuler E. Dementia Praecox oder die Gruppe derSchizophrenien (191 1). (English translation by J Zinkin,New York: International Universities Press, 1950)

19 Schneider K. Primare und sekundare symptome bei derSchizophrenie. Fortschr Neurol Psychiatr 1957;25:487-90

20 Sartorius N, Jablensky A, Shapiro R. Two-year follow-up of patients included in the WHO International PilotStudy of Schizophrenia. Psychol Med 1977;7:529-41

21 Kraepelin E. Die Erscheinungsformen des Irreseins.Zeitschrift der gesamten Psychiatrie und Neurologie1920;62:1-29

(Accepted 19May 1986)

Multicultural studies and the nature of schizophrenia: a review

Documents