Table 3:

Structural covariance network studies on risk developing schizophrenia and related psychotic disorders

Author	Sample size and goals of the study	Field strength/Acquisition scheme/Atlas used/Analysis software used	Methods	Main findings
Liu et al. (2021)	A CHR dataset: 99 high-risk individuals, and 97 HC among 7 other datasets that included first-episode and chronic patients. Goals were to use normative models in stratifying individual patients and to externally validate normative models with symptoms and genotypes	Scan data obtained on either 3T or 1.5T scanners. VBM and CAT12 were used for initial processing AAL2 atlas was used for parcellation.	Used the normative model that was developed in bioinformatics to construct individual-specific gene expression networks. Group level SCN of HC was first built after which one patient was added to create perturbation network. Perturbation networks were developed for each patient. From the Z-scores, individual differential SCNs were built. The three chronic and one CHR datasets were used to explore the stability of our results across different disease stages.	SZ patients were highly heterogenous in their SCNs. Despite high degree of heterogeneity, hippocampus and bilateral putamen/globus pallidus edges grouped patients into two opposing covariance patterns. Among the 4317 edges about 1/3rd edges were significantly different, and only about 28% of these were shared by at least 2 patients. Altered edges correlated with higher hallucinatory scores. Comparison of group level SCN with the IDSCNs showed that the accumulated differences in each patient contributed more strongly. Five variants of 4 genes were associated with edges of the hippocampus and putamen. CHR subjects showed similar group differences as the patients and two subtypes were present even before the onset of the illness.
Pu et al. (2020)	21 subjects with APS and 24 HC. To examine altered cerebello-cerebral structural covariance in APS.	T1 weighted data acquired on a GE 3 T MR scanner using an 8-channel brain-phased array coil. Images preprocessed using the DARTEL on SPM12. Cerebellar regions identified using Spatially Unbiased Infratentorial (SUIT) template and toolbox.	Cerebello-cerebral structural covariance was calculated using the correlation between cerebellar volumes and voxel-wise cerebral GM volume by residualizing for age, gender and ICV. Multiple tests were corrected using familywise error correction. Correlations tests used to study SCNs.	Correlation between the rectal gyrus/medial frontal gyrus was reduced in APS individuals compared to HC. No other covariance was observed to be significant. Correlation of structural correlation with symptom measures were not reported but the correlations with volumes was reported
Heinze et al. (2015)	133 subjects at UHR for psychosis and 65 HC followed for 6–13 years. To examine differences in regional volumes and whole-brain seed-based structural covariance.	1.5 Tesla data acquired at 1.5 mm thickness examined. Images preprocessed using SPM8. GM intensities from 4mm radius spheres placed on the ROIs within the DMN, salience network, executive control networks, visual, auditory, motor, speech, and semantic networks defined on MarsBar toolbox.	Whole-brain patterns of seed-based structural covariance in both hemispheres in each group using TFCE. Statistical threshold for the correlation maps was set at familywise error corrected p<0.01 but no hard threshold-based clustering. Mean GM volume of each seed region and age were covariates. Correlation methods were used to examine the networks.	Higher covariance in motor and executive control networks but lower covariance in DMN in UHR subjects compared to controls. In UHR subjects who converted to psychosis showed alterations in SCN in 4 of the 7 networks examined (motor, executive control, salience, and default mode networks) All these networks were similar to the canonical intrinsic connectivity and structural covariance networks
Shi et al. (2012)	26 neonates at genetic risk for SZ and 26 demographically matched healthy neonates. Goal was to identify inter-regional interactions in high-risk neonates	MRI data acquired on a 3 Tesla head-only scanner. T1 images at 1 mm slices and T2 images at 1.95 mm. DWI data also acquired at 2 mm isotropic voxels. 90 cortical and subcortical regions labeled using the AAL atlas adapted for the neonatal brain space. Fiber tracts were reviewed in ParaView (www.paraview.org).	Graph theoretic methods used for network analysis implemented using the Brain Connectivity Toolbox. Network sparsity, small-world properties such as global efficiency, clustering coefficient and pathlength, connectional distance, modularity and centrality were calculated.	High risk neonates had lower global efficiency, higher local efficiency, clustering co-efficient, pathlength and longer mean Euclidean distance. Male neonates had higher global but reduced local efficiency. High risk neonates had hubs in the frontal, temporal, occipital lobes while control neonates had hubs in the frontal, temporal, parietal, and subcortical regions. High risk neonates had fewer hubs with higher betweenness compared to healthy neonates who had higher number of hubs with lower betweenness. WM network did not show differences in global and local efficiencies and mean connectional distance but larger clustering coefficients.
Bhojraj et al. (2010)	64 adolescent and young adult offspring of schizophrenia/schizoaffective disorder patients and 80 HC.		1.5 Tesla MRI data with 1.5 mm slices were examined for inter-regional correlations of the default mode network (anterior cingulate, inferior parietal, posterior cingulate, medial-prefrontal and lateral temporal neocortices and the precuneus) and the DLPFC. Correlation tests	Inter-regional correlations for two DLPFC default regions (L. DLFPC-left lateral temporal cortex and R. DLPFC-right anterior cingulate cortex) and 7 region pairs in the default regions (bilateral anterior cingulate, bilateral lateral temporal cortex, bilateral posterior cingulate, and right inferior parietal) were significant in offspring. Inter-correlations for none of the region-pairs reached significance in controls.

Abbreviations used: APS, Attenuated Psychosis Syndrome; DARTEL, diffeomorphic anatomical registration through exponentiated Lie algebra; DLPFC, Dorsolateral Prefrontal Cortex; DMN, Default mode network; GM, Gray Matter; HC, Healthy Controls; ICV, Intracranial volume; SCN, Structural Covariance Network; SPM, Statistical Parametric Mapping; SZ, Schizophrenia; UHR, Ultra-high risk;