T211. BASIC SELF-DISTURBANCE AS A PREDICTOR OF DETERIORATION IN ATTENUATED PSYCHOSIS: A 1-YEAR FOLLOW-UP STUDY AMONG COMMUNITY-DWELLING ADOLESCENTS

Abstract Background Phenomenological research indicates that disturbance of the basic sense of self may be a core phenotypic marker of schizophrenia spectrum disorders. Basic self-disturbance (SD) refers to a disruption of the sense of first-person perspective and self-presence that is associated with a variety of anomalous subjective experiences. Recent studies including from our group provided first, preliminary support for the notion that SD is related to attenuated psychosis symptoms (APS) and depression among clinical (i.e., treatment-seeking) and non-clinical samples of non-psychotic adolescents. However, very few studies, if any at all, have looked at the ability of SD to predict change in APS and depression over time. The goal of this study was to address this lacuna in the literature by examining the unique and added contribution of SD to the prediction of change over time in APS and depression among community-dwelling adolescents. Methods The 1-year longitudinal relationship between SD and change in APS and depression were explored in a sample of 100 non-help-seeking adolescents (age 13–15) from the community. SD was assessed with the Examination of Anomalous Self-Experience (EASE), prodromal symptoms and syndromes were assessed with the Structured Interview for Prodromal Syndromes (SIPS), present and lifetime diagnoses of schizophrenia-spectrum and depression disorders were assessed with the Kiddie Schedule for Affective Disorders and Schizophrenia (K-SADS), level of distress with the Mood and Anxiety States Questionnaire (MASQ), and psychosocial functioning with the with the Strength and Difficulties Questionnaire (SDQ). Results Seventy-seven (77%) adolescents out of the 100 that had been assessed at baseline were available and agreed to participate in the 1-year follow-up (Mean=1.4, S.D.=0.8). Except for a diagnosis of an affective disorder, which was slightly less prevalent among those who returned for the follow-up assessment, there were no significant differences between those who were available and those who lost for the follow-up assessment on any of the major socio-demographic or clinical variables at baseline. Consistent with our first hypothesis, SD at baseline predicted a significant amount of variance in APS change over time (R-squared=0.10, F= 8.61, p=0.004). However, inconsistent with our second hypothesis, SD at baseline did not have a significant added contribution to the prediction of APS change when APS at baseline was controlled for (R-squared difference=0.02, F=1.83, p=0.18). Discussion These results provide preliminary support for a prospective association between SD and deterioration in prodromal symptoms among adolescents from the community. However, they fail to support an added value of SD over and above baseline APS for the prediction of APS deterioration. Because SD was assessed only at baseline, they leave unanswered the degree to which change in SD is associated with a change in APS and depression.

Background: Although many people with schizophrenia-spectrum disorders report high levels of social anhedonia, it is not clear what differentiates those patients who self-report social anhedonia from those who do not. Moreover, the extent to which the hedonic functioning of severely disordered patients is associated with their clinical symptoms or with personality-related factors remains unresolved. Methods: We administered the Anticipatory and Consummatory Interpersonal Pleasure Scale (ACIPS; Gooding & Pflum, 2014), a self-report measure designed to assess hedonic capacity for social and interpersonal pleasure, to 125 consecutively admitted inpatients with schizophreniaspectrum disorder. The (81 schizophrenia, 44 schizoaffective disordered) patients were assessed in terms of their illness and symptom severity. They were also administered measures of self-efficacy (GSES; Jerusalem & Schwarzer, 1992), quality of life (Q-LES-Q-18;Ritsner et al., 2005), and recovery level (RAS-20;Salzer, 2010). Based on total ACIPS scores, two cut-off points were defined in order to classify participants as 'normally hedonic', 'hypohedonic' or 'anhedonic'. Results: The ACIPS negatively correlated with 8 PANSS items: conceptual disorganization (P2, r=-0.24, p<0.01), hallucinatory behavior (P3, r=-0.28, p<0.01), suspiciousness (P6, r=-0.31, p<0.001), emotional withdrawal (N2, r=-0.24, p<0.01), stereotyped thinking (N7, r=-0.19, p<0.05), tension (G4, r=-0.23, p<0.01), G5 mannerism and posturing (G5, r=-0.22, p<0.05), and disturbance of volition (G13, r=-0.26, p<0.01).In addition, the ACIPS positively correlated with self-efficacy, self-esteem, perceived social support, subjective quality of life, and recovery scale scores. Discussion: The ACIPS is a reliable and valid means to measure social anhedonia in a clinical sample. The findings revealed that the self-reported hedonic functioning of schizophrenia-spectrum patients is associated with both clinical symptomatology as well as some personality-related variables. Suggestions for further clinical and research applications using the ACIPS will be provided.

University of Haifa; 2 University of Copenhagen
Background: Phenomenological research indicates that disturbance of the basic sense of self may be a core phenotypic marker of schizophrenia spectrum disorders. Basic self-disturbance (SD) refers to a disruption of the sense of first-person perspective and self-presence that is associated with a variety of anomalous subjective experiences. Recent studies including from our group provided first, preliminary support for the notion that SD is related to attenuated psychosis symptoms (APS) and depression among clinical (i.e., treatment-seeking) and non-clinical samples of non-psychotic adolescents. However, very few studies, if any at all, have looked at the ability of SD to predict change in APS and depression over time. The goal of this study was to address this lacuna in the literature by examining the unique and added contribution of SD to the prediction of change over time in APS and depression among community-dwelling adolescents. Methods: The 1-year longitudinal relationship between SD and change in APS and depression were explored in a sample of 100 non-help-seeking adolescents (age 13-15) from the community. SD was assessed with the Examination of Anomalous Self-Experience (EASE), prodromal symptoms and syndromes were assessed with the Structured Interview for Prodromal Syndromes (SIPS), present and lifetime diagnoses of schizophrenia-spectrum and depression disorders were assessed with the Kiddie Schedule for Affective Disorders and Schizophrenia (K-SADS), level of distress with the Mood and Anxiety States Questionnaire (MASQ), and psychosocial functioning with the with the Strength and Difficulties Questionnaire (SDQ). Results: Seventy-seven (77%) adolescents out of the 100 that had been assessed at baseline were available and agreed to participate in the 1-year follow-up (Mean=1.4, S.D.=0.8). Except for a diagnosis of an affective disorder, which was slightly less prevalent among those who returned for the follow-up assessment, there were no significant differences between those who were available and those who lost for the follow-up assessment on any of the major socio-demographic or clinical variables at baseline. Consistent with our first hypothesis, SD at baseline predicted a significant amount of variance in APS change over time (R-squared=0.10, F= 8.61, p=0.004). However, inconsistent with our second hypothesis, SD at baseline did not have a significant added contribution to the prediction of APS change when APS at baseline was controlled for (R-squared difference=0.02, F=1.83, p=0.18). Discussion: These results provide preliminary support for a prospective association between SD and deterioration in prodromal symptoms among adolescents from the community. However, they fail to support an added value of SD over and above baseline APS for the prediction of APS deterioration. Because SD was assessed only at baseline, they leave unanswered the degree to which change in SD is associated with a change in APS and depression.

T212. THE INTRINSIC ORGANIZATION OF SYMPTOMS MARKS TRANSITION FROM HIGH-RISK STATE TO EARLY PSYCHOSIS: A PHENOMENOLOGICAL CONNECTIVITY STUDY
Lena Palaniyappan* ,1 , Tushar Das 1 , Fabienne Harrisberger 2 , Undine E. Lang 2 , Anita Riecher-Rössler 2 , André Schmidt 2 , Stefan Borgwardt 2 1 University of Western Ontario; 2 University of Psychiatric Clinics Basel (UPK) Background: In psychiatric practice, when symptoms "come together" we call the resulting construct as a diagnosis. We believe that there is a disease process that binds together, enabling co-occurrence of varied symptoms. We use either diagnostic or syndromic labels to describe this construct (e.g. positive syndrome, negative syndrome, schizophrenia, at-risk mental state). An emerging idea, promoted by network theorists, is that symptoms may relate by their own intrinsic nature, with no external constructs bringing them together e.g. paranoia leads to social withdrawal, loss of appetite leads to loss of weight etc. This intrinsic organisation of symptom relationships can be studied using network models by applying graph theory to symptom data. Methods: We recruited 63 subjects with at-risk mental state [on the basis of Melbourne PACE criteria] but no transition (ARMS-NT), 16 that later developed psychosis (ARMS-T) and 38 drug-naïve patients with first-episode psychosis (FEP) from Basel, Switzerland. Symptoms were measured using Brief Psychiatric Rating Scale. Clinical symptoms can be construed as a system of individual elements (24 nodes) and their relationship (24x23 possible edges) within a group. We estimate each individual's contribution to the intrinsic organisation of symptoms using a jack-knife bias estimation procedure. Bias values for each pair of symptoms in an individual subject quantified the contribution of that subject to the overall within-group relationship for that symptom pair. Higher values meant greater relationship between the two given nodes in that subject, relative to the rest of the group. We then used Graph Analysis Toolbox, with a range of binarization thresholds based on cost-density of connectivity to extract adjacency matrices.
Results: None of the 24 individual symptoms of BPRS significantly differentiated ARMS-NT from ARMS-T, though a number of symptoms (suspiciousness, hallucinations, disorganisation, motor retardation, hostility and suicidality) showed a gradient of FEP>ARMS-T>ARMS-NT (F test, FDR corrected p<0.05). The small-worldness (F=4.8, p=0.01) and the clustering coefficient (F=10.9, p<0.001) and modularity (F=10.9, p<0.001) of the symptom networks were notably different among the 3 groups, with a gradient of FEP>ARMS-T>ARMS-NT (except for modularity where FEP=ARMS-T). Post-hoc tests revealed significantly high clustering (Hedges's g = 0.60, p<0.05) and high modular organisation (Hedges's g = 0.81, p<0.01) of symptoms in ARMS-T compared to ARMS-NT. There were no differences between ARMS-T and FEP groups. In both ARMS-T and FEP groups, anxiety was the most central symptom. In addition to anxiety, the FEP group also had unusual thought content emerging as a central feature. Discussion: To our knowledge, this is the first study to investigate the intrinsic phenomenological connectivity and its relevance to psychosis in the clinical high-risk population. Risk of transition to psychosis relates to the consolidation of relationship among symptoms (clustering and modularity), but appears unrelated to the severity of symptoms per se. First episode of psychosis could be thought of as a state of high modular clustering among otherwise sparsely connected symptoms. Incongruent clustering (e.g. blunting with anxiety) is reminiscent of Bleuler's concept of ambivalence being a fundamental feature of psychosis. Deconsolidation of symptom clustering could be the key to prevent transition to frank psychosis in high-risk individuals. Reducing the bridging symptoms (esp. anxiety) could weaken the clinical core of a psychotic episode, complementing the pharmacological approaches of reducing dopamine transmission.

T213. THE EFFECT OF ACUTE STRESS ON PARANOID THINKING AND CORTISOL DURING SOCIAL INTERACTION IN HIGH AND LOW SCHIZOTYPES
James Gilleen* ,1 1 Institute of Psychiatry Background: Paranoid thinking, a common symptom of psychosis and schizophrenia, manifests as a sense of threat and may also be indexed by a lack of trust. Stress, in turn, exacerbates psychosis and paranoia, and is a well-established risk factor for schizophrenia as well as a component in several models of psychosis. The present study aimed to determine the impact that acute stress has on paranoid thinking during social interaction in vivo in high (HSZ) vs low (LSZ) schizotypy using an iterated social reciprocity game. The main hypothesis was that HSZ would anticipate grerater social threat and paranoia at baseline compared to LSZ, and moreover that experimentally-induced stress would exacerbate those differences, and thus show that stress differentially modulates how HSZ model the intentions of others. Methods: Matched healthy participants were stratified into HSZ (N=17) and LSZ (N=17) groups and were administered a non-financial, socialreciprocity game against benevolent and malevolent opponents under both stress and no-stress conditions. Stress was manipulated using the MIST (Dedovic et al., 2005) stress paradigm. Cortisol was measured from saliva samples acquired before and after the MIST stress task. Anticipation of threat and trust scores were derived from the social interaction task. Results: At baseline, cortisol levels were not significantly different between HSZ and LSZ but were significantly raised by the stressor task in HSZ (p<.05) but not in LSZ. Higher cortisol at baseline (pre-stress) predicted greater initial and average anticipation of threat (both r=.5, p<.05) (nostress) to other players; and lower initial trust ratings of malevolent, but not benevolent, players. The MIST task significantly elevated stress ratings compared to baseline (p<.001) and following stress, greater change in cortisol from baseline to post-MIST was associated with lower trust ratings