Alcohol Use Disorder and Its Comorbidity With HIV Infection Disrupts Anterior Cingulate Cortex Functional Connectivity

Abstract

Background: Individuals with alcohol use disorder (AUD) have a heightened risk of contracting HIV infection. The effects of these two diseases and their comorbidity on brain structure have been well described, but their effects on brain function have never been investigated at the scale of whole-brain connectomes.

Methods: In contrast with prior studies that restricted analyses to specific brain networks or examined relatively small groups of participants, our analyses are based on whole-brain functional connectomes of 292 participants.

Results: Relative to participants without AUD, the functional connectivity between the anterior cingulate cortex and orbitofrontal cortex was lower for participants with AUD. Compared with participants without AUD+HIV comorbidity, the functional connectivity between the anterior cingulate cortex and hippocampus was lower for the AUD+HIV participants. Compromised connectivity between these pairs was significantly correlated with greater total lifetime alcohol consumption; the effects of total lifetime alcohol consumption on executive functioning were significantly mediated by the functional connectivity between the pairs.

Conclusions: Taken together, our results suggest that the functional connectivity of the anterior cingulate cortex is disrupted in individuals with AUD alone and AUD with HIV infection comorbidity. Moreover, the affected connections are associated with deficits in executive functioning, including heightened impulsiveness.

Keywords: Alcohol use disorder; Alcoholism; Anterior cingulate cortex; HIV infection; Hippocampus; Orbitofrontal cortex; Resting state; fMRI.

Figure 1:

17 functional areas with regions labeled according to SRI24 atlas.

Figure 2:

A) Regression coefficients associated with AUD or AUD+HIV comorbidity diagnosis and their p-values (colorscale threshold: p=0.05/136). The coefficient associated with connectivity between OFC and ACC was significant in AUD, whereas the coefficient associated with the connectivity between the ACC and the HPC area (hippocampus, parahippocampus, amygdala, and temporal poles) was significant in AUD+HIV comorbidity. Four pairs of areas (OFC and Mot, ACC and piC, ACC and HPC, sF and piC, see Figure 1 for definitions) were at trend level after FDR correction for AUD. No other q-value reached FDR-corrected significance. The sign of the coefficients indicates whether the effect was associated with a greater functional correlation (positive coefficient) or a lower functional correlation (negative coefficient). (B) Comparing the direction of these effects with the average connectivity (intercept) indicates whether the factor is associated with a connectivity disruption or a connectivity strengthening. (C) Connectivity differences associated with AUD and HIV (at trend level after FDR correction).

Figure 3:

(A) ACC (green) and OFC (orange). (B) Relation between the logarithm of total lifetime alcohol consumption and the correlation residuals measuring the connectivity without in-scanner motion, acquisition protocol, age, sex, SES and HIV effects between OFC and ACC for all study participants. (C) Effects of total lifetime alcohol consumption on the Stroop color-word reading score mediated by the functional connectivity between OFC and ACC, for all participants, a; b; c are the standard linear regression coefficients (connectivity = a x consumption + α, Stroop = b x connectivity + d x consumption + β, Stroop = c x consumption + γ) reported for mediation analysis. c’=c-ab is the direct effect remaining after removing the mediated effect ab from the total effect c (c’=0 for a full mediation, c’≠0 for partial mediations). The significance of the mediated effect ab was estimated using RMediation [72].

Figure 4:

(A) ACC (green) and HPC (blue). (B) For HIV participants, relation between the logarithm of total lifetime alcohol consumption and the correlation residuals measuring the connectivity without in-scanner motion, acquisition protocol, age, sex, and SES effects for the connectivity between ACC and HPC. For HIV participants, total lifetime alcohol consumption effects mediated by the functional connectivity between ACC and HPC on (C) the Barratt impulsiveness attentional subscale, (D) the Barratt impulsiveness motor subscale, and (E) the entire Barratt Impulsiveness Scale. Total lifetime alcohol consumption was recorded for 84 HIV participants, of whom 25 also had Barratt Impulsiveness Scale scores.