Functional integration in schizophrenia: too little or too much? Preliminary results on fMRI data.
The disconnectivity hypothesis proposes that schizophrenia results from poor or miswired anatomical connections. Theoretically, its functional counterpart should be disintegration. Integration is thought to allow segregated neurons to interact as a coherent whole, referred to as the "core", while the non-interacting part of the brain is referred to as the "rest". In this study, it is suggested that schizophrenia is the result of rest noise interfering with core activity. Two possible causes are assessed: (i) defective core integration, making the core more vulnerable to noise from the rest, or (ii) the rest being too highly integrated, meaning that it can interfere with the core. These hypotheses were tested using fMRI data acquired from 13 stabilized medicated schizophrenic subjects compared to 11 matched controls. Subjects were required to perform a series of lexical decision and retrieval tasks in separate sessions. The brain was divided into 90 components. Integration was defined as the amount of information shared between the components of a sub-system. An iterative aggregation procedure made it possible to identify a core on the basis of the functional clustering index, which assesses the integration of the core relative to its integration with the rest. Correlation of component-pairs within the core was also compared between the two groups. This procedure was repeated for each subject and for each task. Cores did not differ between the two groups, either in terms of integration or in terms of functional clustering index. However, the core was still highly integrated with the rest and the rest was overly integrated in schizophrenic subjects. Both anomalies were correlated with the negative symptoms. These findings were consistent regardless of the task considered. Furthermore, within the core, anterior-posterior correlations were lower in patients (between the frontal and the parietal and posterior cingulate cortices), whereas frontal left-right correlations were excessive. No significant correlation was found with the medication. Thus, it appears that schizophrenia entails a deleterious combination of too much "noisy" integration (from the rest) and too little "significant" integration (anterior-posterior functional connectivity).