High-resolution mapping of alcohol-related brain connectivity in adults using 7T fMRI and multivoxel pattern classification.

Background: Emerging evidence suggests that alcohol use disrupts large-scale brain network interactions, particularly within the triple network model-comprising the Salience Network (SN), Default Mode Network (DMN), and Frontoparietal Network (FPN). However, few studies have examined how these connectivity alterations vary across the full spectrum of alcohol consumption, especially using ultra-high-field imaging and data-driven approaches. This study leverages 7 Tesla resting-state fMRI and multivariate pattern analysis (MVPA) to characterize distinct brain connectivity patterns across heavy, moderate, and non-drinking adults, aiming to identify neural signatures that differentiate alcohol use severity levels.

Methods: We analyzed resting-state functional connectivity data from 69 adults (Mean age - 28.96; SD - 3.49; Range: 22-36) [41M, 28F] drawn from the Human Connectome Project. Participants were stratified into three matched groups (n=23 each): heavy alcohol users (HA), moderate users (MA), and non-users (NA). Alcohol consumption was quantified using the Achenbach Self-Report (ASR) and the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA). Functional connectivity within and between the SN, DMN, and FPN was assessed using both traditional seed-based analyses and MVPA. Between-group differences (HA vs. MA, HA vs. NA, MA vs. NA) were evaluated using Bonferroni corrected statistical tests.

Results: Compared to non-users, alcohol users showed widespread increases in both intra- and inter-network functional coupling. The most striking differences emerged between HA and MA groups, with MVPA revealing unique hyperconnectivity signatures that distinguished these subgroups. Notably, HA individuals demonstrated reduced connectivity between the superior lateral occipital cortex and the precuneus, and hypoconnectivity between the orbitofrontal cortex and language-related regions. No significant sex differences were observed.

Conclusions: This study provides the first evidence from 7T MRI and MVPA that distinct functional connectivity profiles can discriminate levels of alcohol use severity in adults. The observed triple network hyperconnectivity-particularly between heavy and moderate users-may reflect early neurofunctional reorganization or compensatory mechanisms preceding the onset of alcohol use disorder. These findings advance the search for neurobiological markers of risk and resilience along the continuum of alcohol use and underscore the utility of high-field neuroimaging coupled with machine learning in addiction neuroscience.