Table 4 Studies using Polygenic Risk Score to dissect endophenotypes

Nievergelt et al. [68]

Based on data downloaded from the PGC website for BP, MDD and SCZ

940 cases (PTSD), 2554 controls

PRS calculated from GWAS of BD could significantly predict PTSD in U.S. marine soldiers, while PRS from a SCZ and MDD GWAS could not predict PTSD in the U.S. marines.

Middeldorp et al. [69]

13,835 (personality traits)

1) 1738 cases (MDD), 1802 controls. 2) 2101 cases (BP), 3280 controls

Shared polygenic risk factors between neuroticism and MDD and between BP and extraversion. The explained variance of MDD and BP was 0.1%.

Terwisscha van Scheltinga et al. [70]

8690 cases (SCZ), 11,831 controls

152 cases (SCZ), 142 controls

SCZ-PRS was associated with total brain volume (measured by fMRI). PRS was specifically associated with reduced white matter volume, and did not explain variance in grey matter volume. The higher SCZ-PRS the smaller total brain volumes. Disease status was predicted by PRS.

Whalley et al. [71]

Based on data downloaded from the PGC website for BP and MDD

70 cases (unaffected, but at familial risk of mood disorder), 62 controls

Correlation between high polygenic risk for MDD and reduced white matter integrity. No association with BP-PRS and white matter volume.

Walton et al. [72]

3322 cases (SCZ), 3587 controls

255 cases (SCZ)

Increased polygenic risk for SCZ is associated with neural inefficiency in the dorsolateral prefrontal cortex.

Hall et al. [73]

Based on data downloaded from the PGC website for SCZ, BP and MDD

271 cases (SCZ or psychotic BP), 128 controls

There is a genetic overlap between SCZ loci and gamma oscillation and between BP loci and P3 amplitude. Patients with a high SCZ-PRS had reduced gamma response, and patients with a high BP-PRS had smaller P3 amplitude. SCZ-PRS explained 4% of the variance in gamma oscillation phenotype and BP-PRS explained 3% of the variance in P3 amplitude phenotype.

Holmes et al. [74]

9240 cases (MDD), 9519 controls

438 healthy cases

There is a significant association between increasing polygenic burden for MDD and reduced cortical thickness in the left mPFC. MDD-PRS accounts for 4–9% of the phenotypic variance of the amygdala prefrontal cortex thickness.

Whalley et al. [22]

7481 cases (BD), 9250 controls

For genetic information: 87 cases (BD), 71 controls. For fMRI data: 73 cases (BD), 52 controls

High BD-PRS was associated with high activity in limbic regions.