Effective connectivity of the hippocampus can differentiate patients with schizophrenia from healthy controls: a spectral DCM approach

We applied spectral dynamic causal modelling (spDCM; Friston et al., 2014) to analyze the effective connectivity differences between the nodes of three resting state networks (i.e. Default mode network/DMN, Salience network/SAN and Dorsal attention network/DAN) in a dataset of 31 healthy controls (HC) and 25 patients with schizophrenia (SZ), all male. Patients showed increased connectivity from the left hippocampus (LHC) to the dorsal anterior cingulate cortex (DACC), right anterior insula (RAI), left frontal eye fields (LFEF) and the bilateral inferior parietal sulcus (LIPS & RIPS), as well as increased connectivity from the right hippocampus (RHC) to the bilateral anterior insula (LAI & RAI), right frontal eye fields (RFEF) and RIPS. Moreover, negative symptoms predicted the connectivity strengths from the LHC to the DACC, the left inferior parietal sulcus (LIPAR) and the RHC, while positive symptoms predicted the connectivity strengths from the LHC to the LIPAR and from the RHC to the LHC. These results reinforce the crucial role of hippocampus dysconnectivity in SZ pathology and its potential as a biomarker of disease severity.