T133. NEURAL CORRELATES OF EMOTIONAL PROCESSING IN PSYCHOSIS RISK AND ONSET – A SYSTEMATIC REVIEW AND META-ANALYSIS OF FMRI STUDIES

BACKGROUND: Behavioural findings suggest that the emotion processing abnormalities typically observed in established schizophrenia are already present in patients with a first episode of psychosis (FEP). Evidence has been less consistent in people at clinical high risk for psychosis (CHRp). While several studies have reported unaltered behavioural performance on emotion identification in people at CHRp compared to healthy controls, some studies have shown poorer negative emotion recognition. A growing number of functional magnetic resonance (fMRI) studies have investigated brain response to emotion processing to elucidate the mechanisms underlying these processes in FEP patients and CHRp individuals. Despite the marked expansion of this field over the last two decades, to date, no systematic review or meta-analysis has attempted to synthesise the evidence on the neural bases of emotion processing in these groups as potential markers of psychosis vulnerability and expression. METHODS: The PubMed and Ovid MEDLINE databases were searched for published English-language articles applying an emotion processing task during fMRI in a FEP and/or a CHRp sample compared to healthy controls. References of included papers were also screened. For CHRp studies, only those including participants by the basic or attenuated symptom presentation criteria were included. Individual study methodology and results were extracted and systematically reviewed. In addition, at present, statistical parametric mapping contrast maps (‘T-maps’) are being collected from study authors and will be meta-analysed using the Seed-based d Mapping method. The contrasts meta-analysed will be the ones most commonly reported in the studies identified, i.e., of all emotion over comparison conditions and of negative emotion over neutral conditions. These will be meta-analysed separately, as behavioural evidence suggests that emotion recognition performance in these populations might be valence specific. RESULTS: For the systematic review, 4,389 papers were identified through the search. 19 relevant fMRI papers were identified and their references were screened. 17 articles were included after full-text screening. Six out of twelve fMRI studies in the FEP population reported lower brain activation to emotion processing tasks compared to healthy controls. Four articles reported region-specific hyper- and hypoactivations and two studies found no significant difference. Of the seven studies in the CHRp population, one study reported lower brain response to emotion relative to healthy controls, two studies found hyperactivations, one study found region-specific increases and decreases, and two studies reported no significant difference. The most consistent finding across studies was lower amygdala activation in FEP participants (n=6). Conversely, in the CHRp population one article found an increase in amygdala response to emotion with age, consistently with one other article but contrasting with another study showing activity decreases in this region. DISCUSSION: To our knowledge, no previous systematic review or meta-analysis has synthesised the fMRI findings of emotion processing in both people at CHRp and a FEP. The present systematic review shows that while more consistent hypoactivations are found in the FEP population, results are less consistent in CHRp studies. The undergoing meta-analysis will quantitatively synthesise these findings. Elucidating the nature of emotion processing aberrances in early psychosis may help understand the functional changes across both vulnerability and symptom emergence phases and inform molecular investigations into its underlying mechanisms.