Polygenic risk for depression and resting-state functional connectivity of subgenual anterior cingulate cortex in young adults

Abstract

Background: Genetic variants may confer risk for depression by modulating brain structure and function; evidence has underscored the key role of the subgenual anterior cingulate cortex (sgACC) in depression. We sought to examine how the resting-state functional connectivity (rsFC) of the sgACC was associated with polygenic risk for depression in a subclinical population.

Methods: Following published protocols, we computed seed-based whole-brain sgACC rsFC and calculated polygenic risk scores (PRS) using data from healthy young adults from the Human Connectome Project. We performed whole-brain regression against PRS and severity of depression symptoms in a single model for all participants and by sex, controlling for age, sex, race or ethnicity, alcohol use severity, and household income. We evaluated the results at a corrected threshold.

Results: We included data for 717 healthy young adults. We found lower rsFC between the sgACC and the default mode network and frontal regions in association with PRS and lower sgACC-cerebellar rsFC in association with depression severity. We also noted differences by sex in the connectivity correlates of PRS and depression severity. In an additional set of analyses, we observed a significant correlation between PRS and somatic complaints, as well as altered sgACC-somatosensory cortical connectivity in association with the severity of somatic complaints.

Limitations: The current findings should be considered specific to subclinical depression and may not generalize to patients with depressive disorders.

Conclusion: Our findings highlight the pivotal role of distinct sgACC-based networks in the genetic predisposition for depression and the manifestation of depression among young adults with subclinical depression. Distinguishing the risk from severity markers of depression may have implications in developing early and effective treatments for people at risk for depression.

Figure 1

Correlations between polygenic risk scores (PRS) and (A) depression T scores (r = 0.07, p = 0.02) and (B) somatic complaints T scores (r = 0.09, p = 0.002). Pearson partial correlation analyses included 1140 participants (n = 618 females) with age, sex, race or ethnicity, the first principal component for drinking, household income, and ancestry proportions as covariates. Note that residuals are presented here (males = blue crosses, females = orange circles). Solid and dashed lines represent the regressions and 95% confidence intervals, respectively.

Figure 2

Correlation between resting-state functional connectivity (rsFC) of the subgenual anterior cingulate cortex (sgACC) and polygenic risk scores (PRS) and depression T scores for (A) all participants, (B) males, and (C) females. Depression T scores and PRS were modelled together, with age, sex (for all participants), race or ethnicity, the first principal component for drinking, and household income as covariates (results evaluated at voxel p < 0.001, uncorrected, and cluster p < 0.05, corrected for family-wise error). Among all participants, higher PRS were correlated with weaker rsFC between the sgACC and the bilateral superior frontal gyri (SFG) and the bilateral posterior cingulate cortex (PCC) and ventral precuneus (vPCu); higher depression T scores were correlated with weaker rsFC between the sgACC and the right (R) cerebellum (CBL). Among males, higher PRS were correlated with stronger rsFC between the sgACC and left (L) CBL and weaker rsFC between the sgACC L SFG; higher depression T scores were correlated with weaker rsFC between the sgACC with the R CBL and L insula (INS). Among females, higher PRS were correlated with stronger rsFC between the sgACC and the bilateral lingual gyri (LING) and calcarine sulcus (CAL). Colour bars show voxel t and Cohen d values. Nil = no significant findings.

Figure 3

Sex differences in the correlations between resting-state functional connectivity (rsFC) of the subgenual anterior cinculate cortex (sgACC) and polygenic risk scores (PRS), including correlation of (A) PRS and rsFC between the sgACC and the left cerebellum (CBL) IX (males: r = 0.33, p < 0.001; females: r = 0.05, p = 0.4; slope test: t = 3.77, p < 0.001), and of (B) PRS and rsFC between the sgACC and the lingual gyrus (LING) and calcarine sulcus (CAL) (males: r = −0.01, p = 0.9; females: r = 0.26, p < 0.001; slope test: t = 3.72, p < 0.001). Sex differences in the correlations between rsFC of the sgACC and depression T scores, including correlation of (C) depression T scores and rsFC between the sgACC and the left insula (INS) (males: r = −0.25, p < 0.001; females: r = 0.09, p = 0.09; slope test: t = 4.45, p < 0.001), and of (D) depression T scores and rsFC between the sgACC and the right CBL-IV (males: r = −0.30, p < 0.001; females: r = −0.07, p = 0.2; slope test: t = 3.26, p < 0.001). The data points represent residuals (males = blue crosses, females = orange circles). Solid and dashed lines represent the regressions and 95% confidence intervals, respectively. Slope tests adjusted for multiple comparisons (p < 0.01).

Figure 4

Correlation between resting-state functional connectivity (rsFC) of the subgenual anterior cingulate cortex (sgACC) and somatic complaints T score for (A) all participants, (B) males, and (C) females. The somatic complaints T score was modelled together with polygenic risk score (PRS) and depression T score, controlling age, sex (for all participants), race or ethnicity, the first principal component for drinking, and household income as covariates (results evaluated at voxel p < 0.001, uncorrected, and cluster p < 0.05, corrected for family-wise error). Among all participants, higher scores of somatic complaints correlated with weaker rsFC of the sgACC with a cluster in the left (L) postcentral (PoCG) and supramarginal gyri (SMG). Among males, we did not observe any significant clusters. Among females, higher scores of somatic complaints were correlated with lower rsFC of the sgACC with the dorsal precuneus (dPCu). Colour bars show voxel t and Cohen d values. Results for PRS and depression T score re shown in the Appendix 1, Figure S2. Nil = no significant findings.