The effects of polygenic risks for alcohol misuse on negative emotion processing in young adult binge drinkers

Abstract

Extensive research has documented altered emotion processing in binge drinkers. Genetic risks contribute to problem drinking; however, it remains unclear how genetic risks for alcohol misuse affect behavioral and brain responses to negative emotions. We curated data from the Human Connectome Project and identified 97 binge (69 men) and 379 demographically-matched non-binge (142 men) drinkers performing an emotion task during brain imaging. Alcohol use severity was quantified by the first principal component (PC1) identified of principal component analysis of 15 drinking measures. Polygenic risk scores (PRS) for alcohol dependence were computed for all subjects. With published routines and at a corrected threshold, we evaluated how brain responses to matching negative emotional faces vs. geometric shapes associated with PC1 and PRS in a linear regression, with age, sex, and race as covariates. Higher PC1 and PRS were both significantly correlated with greater symptom severity of somatic complaints in binge drinkers. Bingers relative to non-bingers showed stronger activation of a wide array of frontal, parietal and occipital regions and the insula in positive correlation with the PRS. These genetic risk markers featured more prominently in male than in female binge drinkers. In contrast, regional responses in the default mode network may represent sex-shared markers of the genetic risks. These findings suggest altered neurobiological processes of negative emotions in correlation with genetic risks for alcohol misuse in binge drinkers. Longitudinal studies are needed to show how dysfunctional negative emotion processing interlinks the genetic risks and problem drinking.

Fig. 1. Emotion-related brain responses linked to drinking severity and polygenic risk score (PRS) in binge drinkers.

A All, B men, and C women. PC1 and PRS were modeled together in the whole-brain regression, with age, sex (for all subjects), and race as covariates. The results were evaluated at voxel p < 0.001, uncorrected in combination with cluster p < 0.05 family-wise error (FWE) corrected. No clusters showed significant negative correlations. Color bar shows voxel T and Cohen’s d values. L: left; R: right; CBL: cerebellum; FG: fusiform gyrus; LG: lingual gyrus; PHG: parahippocampal gyrus; INS: insula; IFGoper: inferior frontal gyrus pars opercularis; IOG: inferior occipital gyrus; CAL: calcarine sulcus; SOG: superior occipital gyrus; PCC: posterior cingulate cortex; PCu: precuneus; IFGorb: inferior frontal gyrus pars orbitalis; IFGtri: inferior frontal gyrus pars triangularis; MOG: middle occipital gyrus; IPG: inferior parietal gyrus; Nil: no significant findings.

Fig. 2. Regional brain activations to negative emotion processing in correlation with drinking severity (PC1) and polygenic risk score (PRS) in non-binge drinkers.

A all, B men, and C women. PC1 and PRS were modeled together in the whole-brain regression, with age, sex (for all subjects), and race as covariates. The results were evaluated at voxel p < 0.001, uncorrected in combination with cluster p < 0.05 family-wise error (FWE) corrected. No significant negative correlates were identified. Color bar shows voxel T and Cohen’s d values. L: left; LG: lingual gyrus; CAL: calcarine sulcus; SFG: superior frontal gyrus; SMA: supplementary motor area; Nil: no significant findings.

Fig. 3. Group differences between binge (orange circle) and non-binge drinkers (green cross) in the correlations of regional activation to negative emotion processing with drinking PC1 and PRS.

(A) PC1 vs. L CBL-IX; (B) PC1 vs. LG/CAL/SOG; (C) PC1 vs. L SFG; (D) PRS vs. L INS/IFGoper; (E) PRS vs. L CBL-VI/FG/LG; (F) PRS vs. PCC/R PCu; (G) PRS vs. R PHG/FG; (H) PRS vs. L CBL-VI/IOG; and (I) PRS vs. L LG/CAL. Note that (A), (B), and (D) to (H) show the clusters identified from whole-brain regressions in binge drinkers, while (C) and (I) show those identified in non-binge drinkers. The data points represent residuals after controlling for age, sex, and race. Solid and dashed lines represent the regressions and 95% confidence intervals, respectively. The coefficients r’s values for the correlations [***p < 0.05/(9×2) = 0.00278] as well as slope t’s and p’s values for group differences in the correlations are presented. Slope test results in bold for p < 0.05/9 = 0.00556 to correct for multiple comparisons. Note: PRS: polygenic risk score; L: left; R: right; CBL: cerebellum; LG: lingual gyrus; CAL: calcarine sulcus; SOG: superior occipital gyrus; INS: insula; IFGoper: inferior frontal gyrus pars opercularis; FG: fusiform gyrus; PCC: posterior cingulate cortex; PCu: precuneus; PHG: parahippocampal gyrus; IOG: inferior occipital gyrus; SFG: superior frontal gyrus.

Fig. 4. Sex differences in the correlations of regional activation to negative emotion processing with drinking PC1 and PRS in binge drinkers (men: blue circle; women: magenta cross).

(A) PC1 vs. R LG/CAL/SOG; (B) PRS vs. L CBL-VI/FG/LG; (C) PRS vs. L INS/IFGoper; (D) PRS vs. PCC/R PCu; (E) PRS vs. L IPG; (F) PRS vs. R LG/CAL; (G) PRS vs. L CBL-VI/IOG; (H) PRS vs. L MOG; and (I) PRS vs. R IFGorb/IFGtri. The data points represent residuals after controlling for age and race. Solid and dashed lines represent the regressions and 95% confidence intervals, respectively. The coefficients r’s values for the correlations [***p < 0.05/(9 × 2) = 0.000278] as well as slope t’s and p’s values for sex differences in the correlations are presented. Slope test results in bold for p < 0.05/9 = 0.00556 for multiple comparisons. Note: PRS: polygenic risk score; L: left; R: right; LG: lingual gyrus; CAL: calcarine sulcus; SOG: superior occipital gyrus; CBL: cerebellum; FG: fusiform gyrus; INS: insula; IFGoper: inferior frontal gyrus pars opercularis; PCC: posterior cingulate cortex; PCu: precuneus; IPG: inferior parietal gyrus; IOG: inferior occipital gyrus; MOG: middle occipital gyrus; IFGorb/IFGtri: inferior frontal gyrus pars orbitalis and triangularis.