RESUMO
There is conflicting evidence on the efficacy of exercise as intervention for psychosis. This article aims to analyze the effect of exercise on psychotic symptoms. A database search was conducted in PubMed, Web of Science, Scopus, ScienceDirect, EBSCO and Cochrane CENTRAL, based on a protocol (PROSPERO: CRD42022326944). Papers available by March 2023 assessing exercise interventions in psychotic patients were included. A significant improvement was found in Positive and Negative Syndrome Scale (PANSS) positive symptoms (MD = -0.75 [-1.35, -0.15], p = 0.01), with large effect sizes for PANSS-negative and general symptoms (-2.14 [-3.36, -0.92]) and (-2.53 [-3.15, -1.91]), respectively. Heterogeneity was high among studies, 49 and 73% for PANSS-positive and negative symptoms, and low, 0%, for general symptoms. It was hypothesized that functioning of specific brain areas, such as the temporal lobe and hippocampus, may underlie the improvement seen with exercise. Based on neuroimaging/neurophysiology studies, we propose a neurobiological model accounting for the association between exercise and psychotic symptom improvement.