A team of scientists from McLean Hospital and Yale University have been examining the human brain through fMRI scans, to try to determine whether mental health illness may leave physical signs on the brain. The study was published in the Proceedings of the National Academy of Sciences of the United States of America and is available to read through the Creative Commons Licence.
The scientists recruited participants who had already had fMRI scans, using similar scanners. 1010 patients were initially recruited: 210 had been already diagnosed with a primary psychotic disorder, with 137 of these meeting the criteria for schizophrenia or schizoaffective disorder, 73 had been diagnosed with bipolar disorder with psychosis, 192 who had a primary affective disorder without psychosis, 57 were patients seeking treatment for unipolar depression, 109 patients who were not seeking treatment, but had unipolar depression and 608 healthy patients for comparison, who had been recruited through another study, involving similar scanners. Diagnosis was through a standardised test. Comparison patients were matched to the other groups on age, gender, race, left or right handed.
The scientists conducted rest scans, where the patient was asked to lie in the scanner with their eyes open to enable the scientists to gather data about spontaneous fluctuations in the brain. The scientists used connectomics or the ability to measure the brain as a whole rather than concentrating on different parts. This set the study apart, as it was looking for similarities across illnesses rather than focusing on just one. The scientists examined individual systems as well as higher order systems such as the cognitive system would work.
The study found that some of the mental health disorders such as schizophrenia, bipolar disorder and depression showed similar pathologies through the scans, and could have the same underlying causes. Previous studies have found that there could be a significant genetic risk for schizophrenia and bipolar disorder, but this study reveals that one system is disrupted through the severity of illness, whether it was psychosis or depression. Because the study examined patients from a number of sources, the scientists were able to compare the characterisation of the brain system impairments without being troubled by the labels attached to the disorders. They found that broad frontoparietal network impairments were evident, showing a diverse set of cognitive processes which were impaired in multiple disorders. Similar dysfunctions in context processing were also found in patients who had never been medicated who had schizophrenia, early in the illness. This indicated that some of the frontoparietal subnetworks could link to cognitive capacity and psychiatric systems. Age affects functional changes to the brain, which could be linked to parts of psychopathology which are not limited to traditional diagnostic boundaries.
Evidence has mounted that psychotic illness could be linked to changes in how the brain functions. However default network disruption was found in other patients who had bipolar disorder but without psychosis. Impairment in cognitive processes could be linked with multiple disorders. Having observed such variability within diagnoses, data collection would be required to be more high-throughput to enable to link a specific connectome architecture to symptom profiles in patients. The scientists suggest that it may not be possible to identify isolated features of biology of the brain which link to patients of psychiatric illness, but that they might be able to find a number of different pathologies which may lead to a defining fingerprint. In order to achieve this, they would need to recruit patients from a number of diverse clinical settings at different points in their illness and reassess their current approach to extend beyond the conventional clinic-specific or laboratory-specific collection. A limitation to this study is that a number of research groups have been involved to collect the data, which has then been harmonised, but that it can be difficult to collect consistent results in these circumstances. This has limited some of the findings of the study.
Baker, J.T., et al., Functional connectomics of affective and psychotic pathology, Proceedings of the National Academy of Sciences, (PNAS) 2019; 116 (18)