Frontoparietal areas link impairments of large-scale intrinsic brain networks with aberrant fronto-striatal interactions in OCD: a meta-analysis of resting-state functional connectivity

    loading  Checking for direct PDF access through Ovid


HighlightsMeta-analysis of seed-based resting-state fMRI studies in OCD.Altered connectivity was found in default mode, salience and frontoparietal networks.Between-network hypoconnectivity matched the triple network model of dysconnectvity.General dysconnectivity findings support the aberrant fronto-striatal model.Results underline the relevance of frontoparietal regions for OCD pathophysiology.Neuroimaging studies report evidence for two distinct pathophysiological models of obsessive-compulsive disorder (OCD): disrupted fronto-striatal circuits and impaired large-scale fronto-parietal-limbic intrinsic brain networks, defined by functionally connected (FC) infra-slow oscillations in ongoing brain activity. To synthesize this literature and overcome inconsistencies, we conducted a coordinate-based meta-analysis of 18 whole-brain resting-state functional magnetic resonance imaging (fMRI) studies (541 patients, 572 healthy controls) comparing seed-based FC between OCD patients and healthy controls. In patients, the meta-analysis revealed (1) consistent hypoconnectivity within frontoparietal and salience network, and between salience, frontoparietal and default-mode network, and (2) consistent general dysconnectivity (no specific direction of connectivity change) within default-mode and frontoparietal network, as well as between frontoparietal, default-mode, and salience networks. Between-network hypoconnectivity provides evidence for the triple-network model in OCD, while aberrant within-network connectivity of frontoparietal and striatal regions supports reports of aberrant fronto-striatal circuitry. Therefore, results corroborate both models of OCD pathophysiology and link them by underlining the importance of intrinsic connectivity of frontoparietal regions which are common to both models.

    loading  Loading Related Articles