Connectome-based Prediction of Craving Across Substance Use Disorders: A Transdiagnostic Approach

Kurzzusammenfassung in leicht verständlicher Sprache

Craving beschreibt das starke Verlangen nach einer Substanz und ist ein zentrales Problem bei Suchterkrankungen, weil es Rückfälle begünstigt und Therapien erschwert. In dieser Studie wollten wir herausfinden, ob es ein gemeinsames Muster im Gehirn gibt, das Craving bei verschiedenen Süchten vorhersagen kann. Dafür untersuchten wir Gehirndaten von Menschen mit unterschiedlichen Substanzproblemen. Wir fanden heraus, dass ein bestimmtes Netzwerk im Gehirn das Ausmaß des Cravings über verschiedene Süchte hinweg zuverlässig vorhersagen kann. Diese Ergebnisse können helfen, Sucht besser zu verstehen und individuellere Behandlungen zu entwickeln.

Abstract

Hintergrund und Fragestellung

Craving, defined as an intense desire to engage in substance use behavior, is a core feature of substance use disorders and a key predictor of treatment outcomes. However, the neurobiological mechanisms underlying craving remain poorly understood. The aim of the present study was to identify a transdiagnostic functional brain network predictive of self-reported craving across different substance use disorders, and to validate this network in independent datasets.

Methoden

In this cross-sectional study, we applied Connectome-based Predictive Modeling (CPM), a machine-learning approach, to predict the level of self-reported craving based on resting-state functional magnetic resonance imaging (fMRI) data. The sample comprised 78 individuals meeting diagnostic criteria for cannabis, opioid, or tobacco use disorder. Predictive performance was quantified as the correlation between model-estimated and observed craving scores. For external validation, we tested the identified craving network in an independent sample of 41 patients with alcohol dependence, as well as in a second independent sample of 28 individuals with tobacco use disorder assessed during experimentally induced abstinence.

Ergebnisse

Using CPM, we identified a functional brain network that significantly predicted craving across substance use disorders. Key regions within this network included the right medial orbitofrontal cortex, the posterior cingulate cortex, and the lateral prefrontal cortex. In an external sample, the network generalized to alcohol dependence: Positive network connections were primarily associated with cognitive aspects of craving, whereas negative connections were more strongly related to motivational components. Among individuals with tobacco use disorder, the network additionally predicted within-subject changes in craving between baseline and abstinence.

Diskussion und Schlussfolgerung

This study identifies a robust, transdiagnostic neural signature underlying craving in substance use disorders. Its sensitivity to intra-individual fluctuations in craving highlights its potential for personalized approaches in addiction treatment. Future research should examine the extent to which this network can also predict long-term outcomes such as relapse risk and treatment success.

Interessenskonflikte sowie Erklärung zur Finanzierung

I, along with my co-authors, declare that over the past three years there have been no financial interests or personal relationships that could have influenced the work presented in this abstract.


This study was supported by the German Federal Ministry of Education and Research (BMBF; grant numbers 01ZX1909C and 01ZX2209C [SysMedSUDs]), the German Research Foundation (DFG; project ID 402170461 [TRR265]), and the ERA-Net NEURON program (project ID 01EW1112-TRANSALC).