Discussion: Theta activity at occipital sites is the most heritable of themeasured frequency bands in line with previous studies. Suggestivelinkage peaks were observed, including the locus for DTNBP1.

doi:10.1016/j.schres.2010.02.370

Poster 143MEG DOES NOT REVEAL IMPAIRED SENSORY GATINGIN FIRST-EPISODE SCHIZOPHRENIA

Silke Bachmann1, Matthias Weisbrod2, Johannes Schroeder3,Andre Rupp41Dpt. of Psychiatry, University of Halle Halle (Saale) Germany;2Psychiatric Hospital Karlsbad-Langensteinbach/University of HeidelbergHeidelberg Germany; 3Dept. of Psychiatry, University of HeidelbergHeidelberg Germany; 4Dpt. of Neurology, University of HeidelbergHeidelberg Germany

Background: The inablitily to adequately suppress the second oftwo identical stimuli is called sensory gating deficit and can bestudied by recording evoked potentials to auditory stimuli, e.g. theP50 and the N100. It has been looked at as the physiologicalcorrelate of schizophrenia patients' perception of being flooded bysensory impressions. According to the notion that the gating deficitconstitutes a genetic trait, we expected to demonstrate thephenomenon in first-episode schizophrenia patients using MEG.Methods: Eighteen inpatients in remission of their first psychoticepisode and 24 healthy, age- and sex-matched control subjectsparticipated in the study. Diagnoses, psychopathologyandhandednesswere assessed with established instruments. Stimulation was per-formed with the double click paradigm (ISI 500 ms, ITI 9-10 sec).Magnetencephalography (MEG) recordings of 15 patients and 18controls entered further analyses with the software BESA for spatio-temporal source analyses and statistical analyses with MATLAB.Results: Neither P50 nor N100 responses differed statisticallybetween the groups. This means that gating was not impaired inthis first-episode sample of schizophrenia patients.Discussion: Conclusions: These results are not in line with themajority of studies on sensory gating in schizophrenia, however,studies on first-episode patients are scarce. The most likely reasonsfor absence of the gating deficit in our study are patients' first-episode status and atypical antipsychotic medication.

doi:10.1016/j.schres.2010.02.371

Poster 144PREDICTION OF PSYCHOSIS BY MISMATCH NEGATIVITY

Mitja Bodatsch1, Stephan Ruhrmann1, Michael Wagner2, Ralf Müller1,Frauke Schultze-Lutter1,5, Ingo Frommann2, Jürgen Brinkmeyer3,Wolfgang Gaebel3, Wolfgang Meier2, Joachim Klosterkötter1,Anke Brockhaus-Dumke1,41University of Cologne Cologne, NRW, Germany; 2Rheinische Friedrich-Wilhelms University Bonn, NRW, Germany; 3Heinrich Heine UniversityDüsseldorf, NRW, Germany; 4Rheinessen Fachklinik Alzey Alzey, RLP,Germany; 5University of Bern Bern, BE, Switzerland

Background: Prevention of psychosis has become a major task, andalthough current at-risk criteria are associated with a tremendouslyincreased risk for psychosis, results of prediction studies alsodemonstrate a need for further characterization of the individualrisk in order to employ risk adapted preventive measures. It has beensuggested that biological parameters potentially differentiate future

converters from non-converters in the at-risk state and providefurther information about the timing of a future conversion. Impairedpre-attentive sensory information processing is regarded as animportant pathophysiological mechanism contributing to schizo-phrenia. It has been shown that the mismatch negativity (MMN), anauditory event-related potential (ERP) that can be understood as anindex of automatic context-dependent information processing, israther specifically reduced in schizophrenia and is correlated witheveryday functioning. It has been proposed, that primary andsecondary auditory, and potentially frontal cortex contribute to thegeneration of the MMN response. On the molecular level, aninvolvement of the glutamate/n-methyl-d-aspartate (glu/NMDA)system has been suggested. A recent study demonstrated thereduction of the magnetoencephalographic duration MMN in at-riskindividuals compared to healthy controls. The present study aimed toinvestigate the hypothesis that at-risk subjects later converting topsychosis show a deficit of the duration MMN compared to subjectsnot converting in a certain period of time. In addition, the possiblecontribution to the prediction of psychosis was evaluated.Methods:Weused an auditory oddball paradigm in combinationwitha visual vigilance control task. Subjects were recruited in theprospective German Research Network on Schizophrenia (GNRS)study, fulfilled the criteria for a late at-risk state (LARS) at baseline andwere antipsychotic-naive throughout the study. Sixty-two subjectsfulfilled the intake criteria. In case of non-conversion (N=27),the clinical follow-up was at minimum 24 months (46.3±13.1).Converting subjects (N=25) showed an average time to conversion of7.0±7.0 months. Repeated measures ANCOVA was employed toanalyse theMMNrecorded at frontal (F3, Fz, F4) and central electrodes(C3, Cz, C4). A Cox regression model served to evaluate the predictivevalue of the duration MMN amplitude.Results: Subjects with later conversion to psychosis showed signifi-cantly reduced duration MMN amplitudes over the six fronto-centralelectrodes compared to non-converting subjects (F(1)=5.16, p<.05).Based on a stepwise backward Cox regression model (c2=4.77,p<.05; beta=.92±.44; Wald(1)=4.46, p<.05; hazard ratio=2.5,95%-CI: 1.07 – 5.87), four electrodes (Fz, F4, C4, C3) served to create anindividual prognostic score. A median split based on the scoregenerated two risk classes; cumulative hazard rates were 0.34 in class1 and 0.85 in class 2. Kaplan-Meier analysis revealed two significantlydifferent survival curves for the classes (class 1: 20.0, 95%-CI: 16.9 –

23.1; class 2: 13.7, 95%-CI: 10.4 – 17.4; Log-rank test, χ2=5.58, p<.05).Discussion: The present findings demonstrate for the first time thatthe duration MMN is significantly reduced in at-risk subjectsconverting to first-episode schizophrenia within a certain period oftime compared to non-converters. Further analyses indicate thatthis MMN deficit may contribute to the prediction of conversion.Moreover, the findings support the notion that biological para-meters are promising candidates for a more individualized estima-tion of risk in terms of magnitude and time.

doi:10.1016/j.schres.2010.02.372

Poster 145IMPAIRED LONG-TERM POTENTIATION OF THE VISUAL EVOKEDPOTENTIAL IN SCHIZOPHRENIA

Idil Cavus1,2, Robert Guglielmino2, Judith Ford3, Brian Roach3,John Krystal2, Ralitza Gueorguieva2, Daniel Mathalon3

1Pfizer R&D New London, CT, USA; 2Yale University New Haven, CT,USA; 3University of California San Francisco, CA, USA

Background: Long-term potentiation (LTP) is a cellular mechan-ism of synaptic plasticity thought to underlie learning and

Abstracts244