2026 ECNP Congress

PS03-3002

Neurobiological subtypes in alcohol use disorder and their phenotypic and clinical profiles

Anna Zilverstand — L.R. Brucar, E. Rawls · University of Minnesota, Department of Psychiatry, Minneapolis, United States

2026 ECNP Congress

Introduction

Alcohol Use Disorder (AUD) is one of the most prevalent and deadly psychiatric disorders. Globally, alcohol misuse causes approximately three million deaths annually and contributes to more than 5% of the global disease burden [1]. Among those who receive treatment, current interventions have limited success [2,3]. This lack of efficacy has been attributed, in part, to the ‘heterogeneity problem’ - which postulates that distinct causal mechanisms underlie psychiatric disorders in different subsets of individuals [4,5]. We propose that this hinders both the efficacy of current interventions and future treatment development.

Aims

While recent efforts have focused on stratifying psychiatric disorders based on underlying neurobiological markers, such approaches remain largely unexplored in AUD. To close this gap, the current study aims to characterize the neurobiological heterogeneity among individuals with AUD and evaluate their reproducibility and clinical relevance.

Methods

We hypothesized that multiple, distinct neurobiological subtypes exist that underly the neurobiological and clinical heterogeneity observed in AUD. To test this, we included participants withcomplete resting-state functional MRI (rs-fMRI), phenotypic and clinical data from a large data set with high quality neuroimaging data - the Human Connectome Project (N=668; 58% Female; 22% AUD [mild AUD: n=109; moderate-severe AUD: n=38], 40% Female within AUD). We applied a semi-supervised clustering algorithm, Heterogeneity through Discriminative Analysis (HYDRA) [6], to graph theory metrics derived from whole-brain rs-fMRI (two scans of 29 minutes per participant) to delineate potential neurobiological subtypes of AUD. We extensively validated the derived neurobiological subtypes for reproducibility using resampling, permutation testing, and temporal stability, and for clinical relevance using comprehensive assessments of multi-modal behaviours and clinical characteristics.

Results

A significant interaction between Cluster (k) and graph theory metric was identified (F(32) = 587.22, p-FDR<.001, ηp² = 0.81). Post-hoc comparisons indicated that, across all repeated iterations, the 2-cluster solution (k = 2; mean [SD] ARI, 0.59 [0.15]) consistently outperformed all other cluster solutions for each graph theory metric. Among the five different graph metrics, the 2-cluster solution for the graph theory measure participation coefficient demonstrated the highest reliability (mean [SD] ARI, 0.74 [0.03]). Subtype 1 showed a rs-fMRI profile with high participation coefficient in sensory-motor networks, but low participation coefficient in frontoparietal/salience networks. This subtype was clinically characterized by elevated externalizing behaviours. Subtype 2 demonstrated the opposite rs-fMRI profile: it had high participation coefficient in frontoparietal, salience, and default-mode networks, but low participation coefficient in the sensory-motor networks. Clinically, this subtype was primarily characterized by internalizing behaviours. The two subtypes did not differ by AUD symptoms or use characteristics.

Conclusions

We believe that this study is the first to characterize neurobiological heterogeneity through a rs-fMRI based subtyping approach in AUD. Results revealed, as hypothesized, unique and distinct subtypes with complex and interpretable neurobiological profiles, and demonstrable clinical relevance and predictive validity. Findings underscore the value and direct clinical potential of brain-based subtyping to guide the development of personalized and neurobiologically informed targeted interventions for AUD.

References

  1. World Health Organization, 2019. Global Status Report on Alcohol and Health 2018. World Health Organization. 472 p.
  2. Sliedrecht, W., de Waart, R., Witkiewitz, K., Roozen, H.G., 2019. Alcohol use disorder relapse factors: A systematic review. Psychiatry Res. Aug;278:97–115.
  3. Pettorruso, M., Di Lorenzo, G., De Risio, L., Di Carlo, F., d’Andrea, G., Martinotti, G., 2024. Addiction biotypes: a paradigm shift for future treatment strategies? Mol Psychiatry. May;29(5):1450–2.
  4. Feczko, E., Miranda-Dominguez, O., Marr, M., Graham, A.M., Nigg, J.T., Fair, D.A., 2019. The Heterogeneity Problem: Approaches to Identify Psychiatric Subtypes. Trends in Cognitive Sciences. July 1;23(7):584–601.
  5. Brucar, L.R., Feczko, E., Fair, D.A., Zilverstand, A., 2023. Current Approaches in Computational Psychiatry for the Data-Driven Identification of Brain-Based Subtypes. Biological Psychiatry. Apr 15;93(8):704–16.
  6. Varol, E., Sotiras, A., Davatzikos, C., 2017. HYDRA: revealing Heterogeneity of imaging and genetic patterns through a multiple max-margin Discriminative Analysis framework. Neuroimage. Jan 15;145(Pt B):346–64.