115 IV. Diagnosis and Phenomenology IV.A. Symptom clusters and syndromes IV.A.1 SYMPTOM STRUCTURE IN SCHIZOPHRENIA: TWO OR THREE- FACTOR MODELS? S. Dollfus 1, B. Everitt 2 and M. Petit 3 1Centre Hospitalier et Universitaire, Centre Esquirol, Cdte de Nacre, 14000 Caen, France, 2Universit$ de Rouen, Centre Hospitalier du Rouvray, 76301 Sotteville les Rouen, France, 3 Biostatistic department, Institute of Psychiatry, London SE5 8AF, UK. Evidence for a syndromic dichotomy (negative-positive) was produced in a number of studies, although many other authors also found evidence in favour of a third, disorganization subtype, including positive formal thought disorder. These symptomatic dimensions were determined essentially by princi- pal component analyses. Aim. As these methods are primarily exploratory techniques, confirmatory factor analyses (CFA) were used in this study in order to test various models such as the two-syndrome (nega- tive-positive), three-syndrome models (negative--positive-- disorganization) and a four-syndrome model (negative--posi- tive-disorganization-relational) suggested by Peralta et al (1994). Methods. These CFA were carried out in 135 schizophrenic patients on the global ratings of SANS-SAPS (Scale for the Assessment of Negative Symptoms--Scale for the Assessment of Positive Symptoms of Andreasen and Grove, 1986) and on 3 SANS items (inappropriate affect intimacy and closeness, relationship). Results. Three kinds of models (two, three, four-factor models) showed non significant chisquare and high comparative fit indexes (CFI from 0.98 to 1) which indicate that they provide an adequate fit to the data. The disorganization factor included bizarre behaviour and inappropriate affect but not positive formal thought disorder. Discussion. The originality of this study was to display not only one symptome structure in schizophrenia but three kinds of models. These results outline that schizophrenia can be reduced neither to two (negative--positive) nor to three (negative-- positive---disorganization) syndromic dimensions and that dis- organization syndrome is more characterized by bizarre beha- vior and inappropriate affect than positive formal thought disorder. IV.A.2 FACTOR DERIVED SUB-SYNDROMES OF SCHIZOPHRENIA AND FAMILIAL MORBID RISKS A.G. Cardno 1, P.A. Holmans 1, I. Harvey 2, M.B. Williams 2, M.J. Owen 1"3 and P. McGuffin 1 ~ Department of Psychological Medicine, University of Wales College of Medicine, Heath Park, Cardiff, CF4 4XN, U.K. 21nstitute of Psychiatry, London, SE5 8AF,, U.K. 3Department of Medical Genetics, University of Wales College of Medicine, Heath Park, Cardiff, CF4 4XN, U.K. The quantitative trait locus (QTL) approach to linkage and association has the potential to detect genes of small effect in schizophrenia. It assumes a quantitative phenotype which corre- lates with genetic loading for the disorder. Potential candidate phenotypes within the clinical syndrome of schizophrenia include factor analysis-derived sub-syndromes, but the genetic basis of these has not yet been fully evaluated. In this study we performed factor analysis on OPCRIT checklist ratings from 66 patients with RDC schizophrenia. Six substantive factors were found, characterised respectively by: positive formal thought disorder; first rank delusions; first rank halluci- nations; negative symptoms; inappropriate affect/bizarre behav- iour and persecutory delusions. Morbid risk of schizophrenia and other non-affective psychoses was determined in probands' first-degree relatives using the family history approach. Using multiple logistic regression, risk in relatives was best predicted by probands' scores on subsyndromes derived from the inappro- priate affect/bizarre behaviour and positive formal thought disorder factors. IV.A.3 FOUR BEHAVIOURAL SYNDROMES OF SCHIZOPHRENIA C.A. Harvey, D.A. Curson, C. Pantelis, J. Taylor and T.R.E. Barnes Dept of Psychiatry, Charing Cross & Westminster Medical School, Fulham Palace Road, London, U K A complementary approach to defining symptomatic sub- types of schizophrenia is to identify characteristic patterns of 'problem behaviours' associated with the capacity of patients to function in the community. In a large epidemiological survey, 404 patients (age~<60) fulfilling Feighner criteria for schizo- phrenia were identified by key informants and assessed using the MRC Social Behaviour Schedule (SBS) and the Manchester Scale (MS). A principal components analysis was used to extract behavioural syndromes from the SBS data. Four behavi- oural syndromes were identified: 'thought disturbance', 'social withdrawal', 'depressed behaviour' and 'anti-social behaviour'