Atypical use of visuospatial context in schizophrenia, bipolar disorder, and subclinical populations: A meta-analysis

Visual perception in people with psychotic disorders is thought to be minimally influenced by surrounding visual elements (i.e. visuospatial context). Visuospatial context paradigms have unique potential to clarify the neural bases of psychotic disorders because a) the neural mechanisms are well-studied in both animal and human models and b) generalized cognitive deficits are unlikely to explain altered performance. However, the published literature on the subject is conflicting and heterogeneous such that a systematic consolidation and evaluation of the published evidence is needed. We conducted a systematic review and meta-analysis of 46 articles spanning over fifty years of research. Articles included behavioral, fMRI and EEG reports in schizophrenia, bipolar disorder, and subclinical populations. When pooling across all paradigm types, we found little evidence of reduced use of visuospatial context in schizophrenia (Hedges’ g=0.20), and marginal evidence for bipolar disorder (g=0.25). The strongest evidence was observed for altered contrast perception paradigms in schizophrenia (g=0.73). With respect to subclinical populations, we observed immense heterogeneity in populations of interest, individual-difference measures, and study designs. Our meta-analysis provided surprisingly weak evidence for the prevailing view that psychotic disorders are associated with a general reduction in use of visuospatial context. Instead, we observed strongest evidence for a specific alteration in the effect of visuospatial context during contrast perception. We propose altered feedback to primary visual cortex as a potential neural mechanism of this effect.