Major depressive disorder (MDD) has been conceptualized within a neuroimmune metabolic oxidative (NIMETOX) framework. From a systems-biology viewpoint, metabolic and oxidative disturbances implicated in MDD may cross conventional diagnostic boundaries and occur in other severe psychiatric disorders such as schizophrenia (SCZ). The relationships between these metabolic signatures and psychosocial exposures (for example, childhood adversity), resilience, and dimensional measures of psychopathology remain incompletely characterized.
The study aimed to characterize shared and divergent serum metabolic signatures across MDD and SCZ, and to examine associations between identified metabolic modules and childhood adversity, resilience, and dimensional clinical phenotypes (mood, psychotic symptoms, suicidal behaviors).
The investigation included 89 participants: 26 healthy controls (HC), 34 individuals with MDD, and 29 individuals with SCZ. An untargeted serum metabolomics platform identified 1,419 metabolites. From these, the authors selected 38 marker metabolites and integrated them into 11 metabolic modules for systems-level analysis. Group comparisons, nested cross-validated orthogonal partial least squares discriminant analysis (OPLS-DA), and multiple regression models were used to relate modules to categorical diagnoses and dimensional clinical measures.
Thirty-eight metabolites were aggregated into 11 modules representing functional or pathway-related metabolic clusters. The study used nested cross-validated OPLS-DA to test discrimination between groups and multiple regression analyses to evaluate how metabolic modules explain variance in dimensional clinical outcomes. Details on specific metabolite identities and module composition were reported by the authors in the primary manuscript and supplementary material.
Compared with healthy controls, both MDD and SCZ showed several convergent metabolic disturbances:
Reduced mitochondrial lipid bioenergetics, indicating compromised lipid-related energy processes.
Lower metabolic resilience, reflecting diminished capacity to maintain metabolic homeostasis under stress.
Increased sphingolipid stress signaling, consistent with activation of lipid-mediated stress pathways.
Elevated oxidative lipid damage, indicating greater lipid peroxidation.
Signs of renin–angiotensin–aldosterone system (RAAS) stress.
In addition, the study observed progressive increases in polyunsaturated fatty acid (PUFA) membrane remodelling across groups, following the pattern HC → SCZ → MDD. Antioxidant-related measures were reported as selectively reduced in MDD but not in SCZ compared with controls.
When metabolic modules were analyzed against continuous clinical measures, several modules explained substantial variance in symptom dimensions:
These results indicate that the same metabolic architecture can relate to specific clinical dimensions across diagnostic categories.
An integrated OPLS-DA model designed to discriminate healthy controls from the combined patient group (MDD + SCZ) achieved a cross-validated area under the curve (AUC) of 0.946, indicating high separation between HC and cases based on the selected metabolic modules. The authors used nested cross-validation to assess model performance.
A metabolic module reflecting terpenoid exposure biotransformation showed a strong association with psychosis and suicidal behaviors. The authors highlight this module as potentially relevant to clinical features linked to exposure or processing of terpenoid compounds.
The study concludes that MDD and SCZ share an interconnected NIMETOX metabolic architecture, characterized by convergent disturbances in mitochondrial lipid bioenergetics, sphingolipid signaling, oxidative lipid damage, RAAS stress, and PUFA membrane remodelling. While many alterations are transdiagnostic, certain features (for example, selective antioxidant reductions in MDD) and module–clinical associations suggest partly distinct psychosocial–metabolic configurations across clinical dimensions.
The authors propose that these findings support a transdiagnostic NIMETOX systems-biology model linking metabolic dysfunction with dimensional psychopathology and provide a framework for future patient stratification and biomarker validation. The report frames the metabolic modules as candidate targets for further mechanistic study and prospective validation.
The study received ethical approval from the Ethics Committee of the University of Electronic Science and Technology of China (IRB No. 30850). The authors state that all necessary participant consents were obtained. All data produced in the study are available from the authors upon reasonable request. The work was supported by the National Nature Science Foundation of China (W2633246). The manuscript is a preprint and has not been peer reviewed; the authors note this limitation for clinical application.