Neural alterations in substance use disorders: A meta-analysis using activation network mapping across and within...

Despite the growing number of task-based functional magnetic resonance imaging (fMRI) studies on substance use disorders (SUDs), results have been difficult to replicate. Potential reasons for the low reproducibility of neuroimaging findings include a lack of alignment with contemporary conceptualizations of brain functioning. Indeed, meta-analytic approaches often rely on univariate assumptions, failing to recognize the brain's inherently connected and dynamic nature. The current meta-analysis was conducted to address these issues and to identify whether individuals with SUDs show consistent network alterations across task domains.

We first performed a traditional coordinate-based meta-analysis (activation likelihood estimation), followed by a network-based meta-analysis (activation network mapping, ANM) of task-based fMRI studies involving individuals with SUDs. Analyses were conducted both across and within task domains.

A total of 120 fMRI studies were identified, comprising 3113 individuals with SUDs. The traditional approach revealed spatial convergence that was restricted to a small number of brain regions both across and within task domains, and showed low reproducibility. In contrast, the ANM approach identified, with high reproducibility, a functional network encompassing the striatum, thalamus, cingulate cortices, and precuneus across task domains. Notably, while this network was consistently engaged across executive function, negative stimuli, and positive stimuli, the drug cue exposure domain revealed unique alterations within the brain's reward system.

Overall, our findings emphasize the importance of adopting a network-based model of brain functioning to advance our neurobiological understanding of SUDs. They also highlight the unique neural network underpinning drug cue exposure tasks in SUDs, which aligns with the incentive-sensitization theory of SUDs.