Individual atrophy patterns clarify latent neural subtypes in addiction

Addiction is a brain disorder marked by profound inter-individual heterogeneity, a factor that limits the clinical utility of group-level neuroimaging findings. To address this, we constructed a precision neuroimaging framework to capture individual variability. We quantified gray matter volume (GMV) deviations in 464 individuals with five substance-related and addictive disorders using a normative model derived from over 1000 healthy controls. Subsequently, we applied a dimensionality reduction technique to decompose these individual deviations into distinct spatial patterns. We observe a shared pattern involving the insula and prefrontal cortex that reflects common underlying biology across different addictions. In contrast, a pattern centered on the basal ganglia captures the differences between specific addiction types. Importantly, only transdiagnostic factors correlated with clinical measures, whereas the heterogeneity factor did not. By resolving neurobiological heterogeneity into distinct, structural atrophy subtypes, this framework offers insights into understanding structural heterogeneity as a local effect of common networks. Addiction affects the brain differently in every person. This study aimed to map these individualized brain changes. We compared brain scans of 464 people with various addictions against over 1000 healthy individuals to measure personalized patterns of brain shrinkage. We found that addiction involves both shared and highly specific physical brain changes. Recognizing these subtypes helps the public understand that addiction is a biologically diverse condition. Wang et al. use a neuroimaging framework to map individualized gray matter shrinkage across five addictive disorders. They identify distinct structural subtypes, revealing both a shared insula-prefrontal vulnerability and substance-specific basal ganglia changes.