Abstract
The "dopamine hypothesis" of schizophrenia suggests the imbalance of the brain dopamine system plays a crucial role in the development of this disorder. Although this hypothesis has been partly supported by early studies, the brain region-specific abnormalities of different dopamine subsystems, and the influential factors on the heterogeneity of dopamine dysfunction of schizophrenia are still unknown. To address these issues, we carried out random-effect meta-analyses by collecting 49 in vivo PET studies (692 patients and 730 controls). Patients exhibited significant regional and population heterogeneity in D(2/3) receptor availability, primarily manifesting as a moderate increased in the striatum of only drug-off patients (Cohen's d = 0.56, 95%CI [0.16; 0.97]), whereas only drug-on patients showed a large decrease in D(2/3) receptor availability in the thalamus, limbic lobe, substantia nigra, and temporal lobe (Cohen's d < -1.40). Drug-off patients also had a small increase in dopamine synthesis in the striatum (Cohen's d = 0.38, 95%CI[0.06; 0.69]), with no replicable dysfunctions in drug-on patients or other brain regions. D(1) receptor availability and DAT showed high heterogeneity but no consistent significances. Finally, Meta-regression indicated that elevated striatal D(2/3) levels correlated with a higher proportion of males, older patients, longer duration and worse symptoms predicted more severe D(1) deficits in the striatum and prefrontal cortex. This study suggests that different dopamine subsystems in schizophrenia are selectively involved across brain regions; moreover, medication status, sex, age, illness duration and symptom diversity have significant impacts on the dysfunctions of dopamine subsystems in schizophrenia.