Skip to main content
. 2021 Aug 30;26(12):7709–7718. doi: 10.1038/s41380-021-01260-5

Fig. 2. Associations between polygenic risk scores for schizophrenia and global cortical and regional subcortical metrics of human brain structure.

Fig. 2

A Barcharts of variance explained by schizophrenia PRS (R2, y-axis) constructed at each of eight probability thresholds (0.0001 ≥ PSNP ≤ 1, x-axis) for each of nine global mean cortical metrics: CT cortical thickness, Vol grey matter volume, SA surface area, IC intrinsic curvature, LGI local gyrification index, FA fractional anisotropy, MD mean diffusivity, NDI neurite density index, ODI orientation dispersion index. Blue bars indicate negative associations and red bars positive associations; asterisks indicate P values for association after FDR correction: *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001. Polygenic risk scores for schizophrenia were significantly negatively associated with global neurite density index and fractional anisotropy. B Barcharts of variance explained by PRS (R2, y-axis) constructed at each of eight probability thresholds (0.0001 ≥ PSNP ≤ 1, x-axis) for NDI measured at each of seven subcortical regions (colors and asterisks code sign and significance of association as in A). PRS was significantly negatively associated with NDI in thalamus, hippocampus, putamen, and caudate; and significantly positively associated with NDI in pallidum.