Talk from Giulio Pergola - Lieber Institute for Brain Development
Short abstract:
Magnetic Resonance Imaging (MRI) has been extensively used to characterize schizophrenia (SCZ) pathophysiology over the past decades. Research has highlighted differences in structural and functional brain characteristics compared to neurotypical controls (NC) both in cortical and subcortical regions, including the hippocampus and the thalamus. However, these measures have yet to be used as potential biomarkers of SCZ, as concerns have been raised regarding their reliability, heritability, and the influence of confounding factors.
We addressed a well-known structural MRI confounding factor, in-scanner head motion. In this study, we analyzed three independent cohorts, including 330 SCZ patients and 1,149 NC. Motion estimates extracted during functional scans used to correct group comparisons in sMRI resulted in reductions in the extent of observed group differences ranging from 14% to 44% across cohorts. Similar results were also found when comparing NC with patients with bipolar disorder (BD), suggesting that sMRI group differences in psychiatric populations may be confounded by head motion and that reported morphometric differences may be overestimated.
We thus focused on brain characteristics related to functional connectivity (FC). We detected stable and heritable connectivity phenotypes by leveraging multiple fMRI scans, including resting state, task-based activations during working and episodic memory, and facial emotion identification. We examined brain FC in 7,431 NC from childhood to adulthood, 201 siblings (SIB) of SCZ patients, 195 SCZ patients, and 1,409 individuals displaying subthreshold psychotic symptoms (PSY). Among the younger SIB, we observed early FC alterations in circuits involving the prefrontal-sensorimotor and cerebellar-occipitoparietal regions that were associated with the polygenic risk score for SCZ. Similar alterations were identified in SCZ patients and young PSY individuals. All findings were pFDR<0.05. These risk-associated FC patterns in younger individuals resembled those observed in older NC, suggesting that as age-related changes in FC progressed, patterns in young SIB gradually approached those seen in NC.
Our findings indicate that while caution should be used when interpreting structural SCZ-related group differences, age-related FC patterns linked to both genetic and clinical risk for SCZ can be reliably detected through fMRI. Age-related FC patterns are also associated with the polygenic risk for SCZ, potentially representing the systemic context of multiple molecular pathways where the genetic risk can be parsed. The next challenge is to define biologically informed polygenic risk to investigate subtypes of genetic SCZ risk and their fMRI correlates. We showcase one example of this approach. This work is expected to boost early detection and develop personalized intervention approaches for SCZ based on genetic etiology.