Microbiota Dysbiosis and the Gut-Brain Axis in Psychiatric Disorders: Molecular Mechanisms, Clinical Evidence, and Therapeutic Perspectives

The gut-brain axis provides a biologically plausible framework for understanding the bidirectional interactions among the intestinal microbiota, immune-inflammatory signaling, neuroendocrine regulation, neurotransmitter metabolism, and psychiatric symptoms. Although numerous recent reviews have examined individual psychiatric disorders, microbial composition, or microbiota-targeted therapies, the field still lacks a sufficiently integrated appraisal that clearly distinguishes associative findings from human studies, mechanistic evidence from preclinical research, and preliminary clinical interventions. This structured narrative review synthesizes clinical, translational, and preclinical studies across major depressive disorder, bipolar disorder, anxiety disorders, schizophrenia, and trauma-related conditions. The main literature search covered publications from January 2015 to January 2026; earlier foundational mechanistic studies of key relevance to gut–brain communication were also considered. Recurrent pathways include altered short-chain fatty acid production, tryptophan-kynurenine metabolism, immune-inflammatory activation, hypothalamic-pituitary-adrenal axis dysregulation, neurotransmitter-related signaling, impaired intestinal and blood-brain barrier integrity, and vagus nerve-mediated communication. Human studies predominantly demonstrate associations between microbial signatures and psychiatric phenotypes, symptom severity, inflammatory status, or treatment response, whereas preclinical models provide stronger mechanistic evidence that microbiota manipulation can influence behavioral, neurochemical, immune, and neuroendocrine outcomes. Probiotics, prebiotics, dietary interventions, and fecal microbiota transplantation show therapeutic potential; however, the current evidence is limited by small sample sizes, methodological heterogeneity, insufficient standardization, and a lack of long-term validation. The distinctive contribution of this review is its structured comparison of three levels of evidence—human observational findings, experimental mechanistic studies, and emerging clinical interventions—thereby clarifying what has been consistently replicated, what remains mechanistically plausible, and what is not yet ready for routine application in psychiatric practice.