Dissociation and Disconnection in Psychiatry and Neurology: A Predictive Coding and Allostatic Framework—A Transdiagnostic Perspective

Dissociation and disconnection are central but often insufficiently integrated concepts in psychiatry and neurology. We propose that these processes represent a transdiagnostic dimension across several psychiatric and neurological disorders, requiring a unified biopsychosocial model. We first examine association as a fundamental principle of life—from cellular to social levels—introducing allostasis as the dynamic context in which association occurs. Dissociation is then analyzed as a process of rupture triggered by traumatic stress, emphasizing cognitive appraisal and cultural context. The predictive coding model is presented as a unifying framework in which dissociation reflects disrupted precision of predictions and prediction errors, particularly regarding interoceptive signals. We distinguish dissociation (phenomenological experience) from disconnection (neural substrate) and disintegration (neurodegenerative/vascular breakdown), and illustrate the model’s application to schizophrenia, anxiety, chronic pain, dementias, parkinsonism, and vascular pathologies. We also discuss the dimensional structure of dissociative phenomena. Finally, we outline therapeutic strategies aimed at restoring association—neuroplasticity, learning, behavioral activation, pharmacological interventions, somatic approaches, and social integration. We conclude that an integrative view of dissociation and disconnection opens new horizons for diagnosis and therapy within the biopsychosocial continuum.