T115. REASONING BIAS, WORKING MEMORY PERFORMANCE, AND A TRANSDIAGNOSTIC PHENOTYPE OF AFFECTIVE DISTURBANCES AND PSYCHOTIC EXPERIENCES IN THE GENERAL POPULATION

Abstract Background There is robust evidence that reasoning biases such as a tendency of jumping to conclusions (JTC) as well as cognitive deficits are associated with psychosis, but evidence on their association with affective disturbances remains inconclusive. Recent findings also suggest a transdiagnostic phenotype of co-occurring affective disturbances and psychotic experiences. This study aimed to investigate whether JTC bias and decreased working memory performance are associated with the co-occurrence of affective disturbances, psychotic experiences (PEs), and psychosis-related help-seeking behaviour (HS) in the general population. Methods The second Netherlands Mental Health Survey and Incidence Study (NEMESIS-2). Trained interviewers administered the Composite International Diagnostic Interview (CIDI) at three time points in a representative general population sample, with N=4.596 individuals who completed all assessments. The beads task and digit-span task were completed to assess JTC bias and working memory performance, respectively. CIDI was used to measure affective disturbances (i.e. depression, anxiety, mania) and an add-on instrument to measure PEs and HS. Results We found that, compared to individuals with neither affective disturbances nor PEs, JTC bias was more likely to be present in individuals with co-occurring affective disturbances, PEs, and HS (moderate psychosis [1–2 PEs]: relative risk ratio [RRR]=1.23, 95% CI 1.03 – 1.48, p=0.023; high psychosis [3 or more PEs or HS]: RRR=1.66, 95% CI 1.26 – 2.19, p<0.001) in models adjusted for socio-demographic characteristics and socio-environmental factors. However, when we additionally adjusted for working memory performance this association was attenuated (moderate psychosis: RRR=1.17, 95% CI 0.98 – 1.41, p=0.088; high psychosis: RRR=1.57, 95% CI 1.19 – 2.08, p=0.002). In line with previous findings, there was no evidence that JTC bias was more likely to occur in individuals with sole presence of affective disturbances (RRR=1.03, 95% CI 0.94–1.13, p=0.492). Further, there was some evidence of a dose-response relationship, as JTC bias was progressively more likely to occur in individuals with affective disturbances as the level of PEs increased or HS was reported (high vs. moderate psychosis: p=0.052). In contrast, compared to individuals with neither affective disturbances nor PEs, a decreased working memory performance was evident in all groups (i.e., affective disturbances only: RRR=0.94, 95% CI 0.90–0.98, p=0.006; PEs only: RRR=0.79, 95% CI 0.69–0.91, p=0.001; co-occurring affective disturbances and moderate psychosis: RRR=0.83, 95% CI 0.75–0.91, p<0.001; co-occurring affective disturbances and high psychosis: RRR=0.76, 95% CI 0.65–0.89, p=0.001). Discussion Our findings suggest that JTC bias and decreased working memory performance may contribute to a transdiagnostic phenotype of co-occurring affective disturbances and PEs. Further, findings support the notion that JTC bias may be specifically associated with psychosis, including in those presenting a transdiagnostic phenotype, while a lowered working memory performance may represent a more broadly distributed vulnerability factor across various symptom domains. Overall, results point to the need to further investigate whether established mechanism and risk factors, described to be involved in the development and maintenance of psychosis, extend to transdiagnostic phenotypes to further corroborate proposed aetiological models and overcome shortcomings of focussing only on specific domains of mental health.


T114. SCHIZOTYPAL PERSONALITY QUESTIONNAIRE-BRIEF: FACTOR STRUCTURE ANALYSIS IN A NONCLINICAL ROMANIAN SAMPLE
Aziz Ferchiou* ,1 , Maria Ladea 2 , Mihai Bran 2 , Marion Leboyer 3 , Ruxandra Slavu 4 , Franck Schurhoff 3 , Andrei Szoke 5 1 INSERM; 2 University of Medicine and Pharmacy "Carol Davila";3 H Mondor Hospital,Fondation FondaMental;4 Clinical Hospital of Psychiatry;5 H Mondor Hospital,Fondation FondaMental Background: The study of schizotypal traits in the general population has been proposed as a way to understand aetiology and pathophysiology of schizophrenia. Self-report measures of psychometric schizotypy like the Brief version of the Schizotypal Personality Questionnaire (SPQ-B) have been shown to be valid, inexpensive and non-invasive tools. Few studies used a Likert-type scale format, which could be better able to allow partial endorsement and to detect more defended respondents than the forced choice format. At our knowledge, no studies of the SPQ-B used validity and social desirability items, to assess the potential impact of random or biased answers.

Methods:
We examined factor structure and internal reliability of a Romanian version of the Schizotypal Personality Questionnaire-Brief (SPQ-B), in a Likert format in a sample of 580 students of Universities of Bucarest, Craiova and Brasov, in Romania. 3 validity items and 5 items of social desirability were added to the 22-items SPQ-B. We investigated the dimensional structure of the Romanian version of the SPQ-B first in the entire sample, and then after eliminating "bad" responders (i.e. those with aberrant answers on the validity items). We used a Principal Components Analysis (PCA) followed by a promax rotation. Factor selection was based on Eigenvalues over 1.0 (Kaiser's criterion), Cattell's scree plot test, and interpretability of the factors. We calculated Cronbach's Alpha for total SPQ-B and for each dimension. Results: Our sample was constituted of 433 women and 147 men. The mean age was 25.5 ± 4.5 years. SPQ-B Likert total scores ranged from 23 to 90 points (mean = 55 ± 12). 71 participants were excluded after taking account of validity questions. Factor analysis of the entire sample resulted in a 3-factor solution that explained 43.8% of the variance. Factor 1 (Cognitiveperceptual; 10 items) includes items related to "ideas of reference", "magical thinking", "unusual perceptual experiences" and "suspiciousness". Factor 2 (Interpersonal; 5 items) includes items related to "social anxiety", "no close friends", and "constricted affect". Factor 3 (Disorganized; 7 items) includes items related to "odd behavior" and "odd speech". Coefficient Alpha for the three subscales and total scale, respectively, were 0.74, 0.78, 0.78 and 0.86. There were no significant differences when the analyses were conducted in the sample of 509 "good" responders' students. Discussion: Factor analysis of the Romanian version of the SPQ-B in a Likert format confirmed the three-factor structure of schizotypy. The SPQ-B and its subscales demonstrated good internal reliability. The use of items of validity and social desirability did not change significantly the results.

T115. REASONING BIAS, WORKING MEMORY PERFORMANCE, AND A TRANSDIAGNOSTIC PHENOTYPE OF AFFECTIVE DISTURBANCES AND PSYCHOTIC EXPERIENCES IN THE GENERAL POPULATION
Christian Rauschenberg* ,1 , Ulrich Reininghaus 1 , Margreet ten Have 2 , Ron de Graaf 2 , Saskia van Dorsselaer 2 , Nicole Gunther 1 , Lotta-Katrin Pries 1 , Sinan Guloksuz 1 , Rajiv Radhakrishnan 3 , Maarten Bak 1 , Jim van Os 4 1 Maastricht University; 2 Trimbos Institute; 3 Yale University School of Medicine; 4 University Medical Center Utrecht Background: There is robust evidence that reasoning biases such as a tendency of jumping to conclusions (JTC) as well as cognitive deficits are associated with psychosis, but evidence on their association with affective disturbances remains inconclusive. Recent findings also suggest a transdiagnostic phenotype of co-occurring affective disturbances and psychotic experiences. This study aimed to investigate whether JTC bias and decreased working memory performance are associated with the co-occurrence of affective disturbances, psychotic experiences (PEs), and psychosis-related help-seeking behaviour (HS) in the general population. Methods: The second Netherlands Mental Health Survey and Incidence Study (NEMESIS-2). Trained interviewers administered the Composite International Diagnostic Interview (CIDI) at three time points in a representative general population sample, with N=4.596 individuals who completed all assessments. The beads task and digit-span task were completed to assess JTC bias and working memory performance, respectively. CIDI was used to measure affective disturbances (i.e. depression, anxiety, mania) and an add-on instrument to measure PEs and HS. Results: We found that, compared to individuals with neither affective disturbances nor PEs, JTC bias was more likely to be present in individuals with co-occurring affective disturbances, PEs, and HS (moderate psychosis [1-2 PEs]: relative risk ratio [RRR]=1.23, 95% CI 1.03 -1.48, p=0.023; high psychosis [3 or more PEs or HS]: RRR=1.66, 95% CI 1.26 -2.19, p<0.001) in models adjusted for socio-demographic characteristics and socio-environmental factors. However, when we additionally adjusted for working memory performance this association was attenuated (moderate psychosis: RRR=1.17, 95% CI 0.98 -1.41, p=0.088; high psychosis: RRR=1.57, 95% CI 1.19 -2.08, p=0.002). In line with previous findings, there was no evidence that JTC bias was more likely to occur in individuals with sole presence of affective disturbances (RRR=1.03, 95% CI 0.94-1.13, p=0.492). Further, there was some evidence of a dose-response relationship, as JTC bias was progressively more likely to occur in individuals with affective disturbances as the level of PEs increased or HS was reported (high vs. moderate psychosis: p=0.052). In contrast, compared to individuals with neither affective disturbances nor PEs, a decreased working memory performance was evident in all groups (i.e., affective disturbances only: RRR=0.94, 95% CI 0.90-0.98, p=0.006; PEs only: RRR=0.79, 95% CI 0.69-0.91, p=0.001; co-occurring affective disturbances and moderate psychosis: RRR=0.83, 95% CI 0.75-0.91, p<0.001; co-occurring affective disturbances and high psychosis: RRR=0.76, 95% CI 0.65-0.89, p=0.001). Discussion: Our findings suggest that JTC bias and decreased working memory performance may contribute to a transdiagnostic phenotype of co-occurring affective disturbances and PEs. Further, findings support the notion that JTC bias may be specifically associated with psychosis, including in those presenting a transdiagnostic phenotype, while a lowered working memory performance may represent a more broadly distributed vulnerability factor across various symptom domains. Overall, results point to the need to further investigate whether established mechanism and risk factors, described to be involved in the development and maintenance of psychosis, extend to transdiagnostic phenotypes to further corroborate proposed aetiological models and overcome shortcomings of focussing only on specific domains of mental health. The association between caffeine consumption and various psychiatric manifestations has long been observed. Methods: We present two cases that show the ability of caffeine to induce psychotic and manic symptoms, and we also review the extant literature on caffeine-induced psychiatric manifestations. Results: On the basis of our own and others' findings, we suggest that caffeine may be related to not only de-novo psychotic or mood symptoms but also to aggravation of pre-existing psychotic or mood disorders. Discussion: We therefore suggest that caffeine consumption among patients with mood or psychotic symptoms should be assessed carefully in clinical practice as part of routine psychiatric evaluations.

T117. INVESTIGATION OF FORMAL THOUGHT DISORDER AND RESPONSE TO TREATMENT IN SCHIZOPHRENIA
Fernando Rocha Loures Malinowski 1 , Bruno Bertolucci* ,1 , Cristiano Noto 1 , Deyvis Rocha 1 , Cinthia Higuchi 1 , Rodrigo Bressan 1 , Ary Gadelha 1 1 Federal University of Sao Paulo (UNIFESP) Background: Formal thought disorder (FTD) is a multidimensional dysfunction characterized by inability to maintain a coherent speech in spoken or written language, poor social cognition and disorganized thought itself. Presence of formal thought disorder has been associated with poor prognosis in schizophrenia, but the association with treatment response is yet to be determinate. Formal thought disorder has a close relation to disorganized symptoms in schizophrenia, which were independently associated with treatment resistance and poor response to standard antipsychotics. Formal thought disorder investigation could provide a clinical construct better delimited to assess disorganized symptoms in schizophrenia. We investigated the association between FTD, remission and treatment resistance in patients with schizophrenia. Methods: This study reunite a sample of 213 patients, between 14 and 69 years, who met DSM-IV criteria for schizophrenia. The analyses were conducted in two samples conducted independently. In both samples, Diagnostic evaluation was performed with the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I), response to treatment was primarily assessed through PANSS, functional impairment was assessed by GAF and disease severity, by CGI. The first sample was a follow-up study that enrolled inpatients. Participants were rated at baseline and after four weeks of antipsychotic treatment. If the participant did not reduce a minimum of 40% of baseline PANSS, the antipsychotic was switched. If the participant did not reduce a minimum of 40% in total PANSS in the following antipsychotic trial, the participant was considered as treatment resistant schizophrenia (TRS) and clozapine, introduced. The second sample was enrolled in an outpatient clinic specialized in schizophrenia. Illness remission was defined as a severity of mild (score of 3 on a scale of 1 to 7) or less for P1, P2, P3, G9, G5, N1, N4 and N6 PANSS`s items. To stablish FTD severity, PANSS items related to high scores at the Thought and Language Index (TLI) were considered: P2, P6, N1, N2, N5, N6, G7 and G9. Results: Most subjects were male (56.8%) and the mean age was 34.42 (±12.33 SD). The FTD failed to associate with remission (t = 4.007, p = 0,491) or treatment resistance (t = -3.768, p = 0.988) in both samples. FTD had a negative correlation with GAF (r = -0.473, p<0.01) and a positive correlation with CGI (r = 0.530, p<0.01). Discussion: FTD had a stronger association with global functioning and severity measures, rather than treatment symptomatic outcomes. In future studies, we will investigate whether FTD show distinctive clinical features commonly related to disorganized syndrome, i.e. earlier age of onset.

T118. IMPACT OF DYSFUNCTIONAL METACOGNITIONS AND WORRY ON DEVELOPMENT OF PARANOIA: A 1-YEAR LONGITUDINAL STUDY IN A NON-CLINICAL SAMPLE
Xiaoqi Sun* ,1 , Suzanne So 1 , Raymond Chan 2 , Chui-de Chiu 1 , Patrick Leung 1 1 The Chinese University of Hong Kong; 2 Institute of Psychology, Chinese Academy of Sciences Background: A worry thinking style has been identified as one of the proximal causal factors for paranoia (Freeman & Garety, 2014). This argument has been supported by the finding that patients with paranoia worry as much as patients with generalized anxiety disorder, and that worry predicts paranoia in non-clinical individuals. Wells (1995) argued that it is when metacognitions about worry (i.e. beliefs about worry and meta-worry) exaggerate worrying that anxiety disorders emerge. It was not clear how metacognitions interact with trait worry in the development of non-clinical paranoia.