Fig. 4: Effects of childhood urbanicity and polygenic risk for depression on stress-associated mPFC function during working memory manipulation.

A In the discovery sample, peak effects in subjects with urban childhoods where polygenic risk for depression correlated with relatively deleterious reduced engagement of mPFC under stress vs. less stress (p < 0.001 uncorrected and p < 0.05 voxel-wise FWE corrected for multiple comparisons within the mPFC ROI). B Plot of the gene–environment interaction between rural or urban childhoods and polygenic risk for depression on stress-related mPFC engagement (p < 0.001 uncorrected and p < 0.05 voxel-wise FWE corrected for multiple comparisons within the mPFC ROI). In the group with urban childhoods, increased polygenic risk for depression resulted in disproportionately suppressed medial prefrontal cortex engagement during stress. This effect was less apparent in those with rural childhoods. C Similar effects were observed in the replication sample, where those with urban childhoods had more deleterious engagement of stress-related mPFC in relation to polygenic risk for depression (p < 0.001 uncorrected and p < 0.05 voxel-wise FWE corrected for multiple comparisons within the mPFC ROI). D These effects were absent in those with rural childhoods, resulting in a gene–environment interaction (p < 0.005 uncorrected). E In the combined sample, polygenic risk for depression was associated with deleterious stress related mPFC engagement (p < 0.001 uncorrected and p < 0.05 voxel-wise FWE corrected for multiple comparisons within the mPFC ROI) in those with urban childhoods. F The absence of these effects in those with rural childhoods resulted in a significant gene–environment interaction (p < 0.001 uncorrected and p < 0.05 voxel-wise FWE corrected for multiple comparisons within the mPFC ROI) in the combined sample.