Translational evidence for habenula's role in human addiction: A scoping review

Addiction is characterized by persistent substance use or engagement in behaviors despite adverse consequences and impaired control, but its expression varies substantially across individuals, substances, and severity. Within one influential neurocircuitry framework, addiction processes are organized into binge, withdrawal, and preoccupation stages. Preclinical research implicates the habenula across these processes but indicates that its contribution varies across drugs, subdivisions, and experimental models. Research on the human habenula in addiction remains fragmented, and the extent to which it converges with preclinical mechanisms is unclear. This scoping review synthesized evidence on the human habenula in addiction and examined its convergence with preclinical findings. Following PRISMA-ScR guidelines, databases and grey literature were searched in November 2025, with literature surveillance continued through April 2026. Human studies reported reduced fronto-habenula white-matter integrity, elevated habenula-striatal resting-state functional connectivity, withdrawal-related habenula-frontal connectivity, increased habenula blood flow during opioid withdrawal, altered task-evoked responses to reward and feedback, and nicotine-related genetic signals. By contrast, morphological studies found no consistent differences in habenula structure across addiction samples. Most human neuroimaging studies could not distinguish the medial and lateral habenula, limiting direct comparison with subdivision-specific preclinical mechanisms. Overall, across human and preclinical studies, the evidence does not support a unitary habenular alteration across addiction. Cross-species convergence was strongest for withdrawal-related activity, particularly for opioids, and for nicotine-related molecular mechanisms, but remained limited by differences in substances, circuits, measures, and paradigms. Future studies using subdivision-sensitive imaging and more comparable cross-species designs are needed.