Daniele Cavaleri1*
Carlo Bassetti1
Giorgio Cucchi1
Pasquale De Fazio2
Renato de Filippis2
Umberto Albert3
Luca Pellegrini3
Giuseppe Carrà1
Francesco Bartoli1- 1School of Medicine and Surgery, University of Milano-Bicocca, Monza, Italy
- 2Department of Health Sciences, University “Magna Graecia” of Catanzaro, Catanzaro, Italy
- 3Department of Medicine, Surgery and Health Sciences, University of Trieste, Trieste, Italy
Mental disorders remain diagnosed primarily through symptom-based classification systems that overlook biological heterogeneity, preventing the identification of mechanistically distinct patient subgroups and precluding pathophysiology-guided treatment selection. Metabolomics offers a promising pathway towards precision psychiatry by capturing dynamic biochemical readouts at the functional endpoint of the omics cascade, integrating genetic, environmental, and pharmacological influences on cellular metabolism. Over the past 15 years, untargeted and targeted metabolomics studies using nuclear magnetic resonance spectroscopy and mass spectrometry have identified consistent patterns of metabolic dysregulation across psychiatric disorders, particularly involving amino acid metabolism, lipid signaling, energy homeostasis, and oxidative stress pathways. Schizophrenia presents disruptions in arginine and proline metabolism, glutathione metabolism, and energy-related processes. Bipolar disorder shows perturbations in branched-chain and aromatic amino acids, kynurenine pathway, and tricarboxylic acid cycle dysfunction with phase-specific metabolic signatures. Major depressive disorder exhibits widespread alterations in amino acid turnover, bioenergetic processes, membrane lipid homeostasis, and glutamate-GABA cycling, with treatment-responsive metabolic changes. Despite these advances, substantial challenges remain: heterogeneous findings with disorder overlap, limited replication cohorts, predominance of cross-sectional designs, confounding by medication and lifestyle factors, pre-analytical variability, and high-dimensional data complexity. Future research requires harmonized multi-site protocols, longitudinal validation studies, multi-platform analytical approaches, integration with genomics, proteomics, and digital phenotyping, and implementation of artificial intelligence frameworks to enhance phenotype discrimination and predictive accuracy. In this mini-review, we provide an overview of current methodologies, major findings, strengths, challenges, and emerging directions in psychiatric metabolomics, with the goal of facilitating the translation of metabolomic insights into clinically applicable, personalized psychiatric treatment.
1 Background
Mental disorders encompass a wide spectrum of heterogenous conditions whose etiologies involve biological, environmental, and psychosocial factors (1). Despite major advances in neuroscience, clinical psychiatry keeps relying on symptom-based diagnostic systems that categorize patients by clusters of observable behaviors and self-reported symptoms but overlook biological heterogeneity, contributing to diagnostic uncertainty and unsatisfactory clinical management (2–4).
The concept of precision psychiatry aims to move beyond the “one-size-fits-all” approach by integrating multidimensional data – including omics, neuroimaging, and digital phenotyping – to stratify patients into biologically meaningful subgroups, predict treatment response, and monitor disease trajectories (5). In this perspective, metabolomics holds particular promise. Metabolites indeed represent the end products of gene-environment interactions and reflect real-time biochemical activity across interconnected pathways (6). Metabolomics thus sits at the downstream end of the omics cascade and provides a sensitive functional readout of cellular and systemic metabolism and its changes in response to physiological and pathological states, environmental exposures, and pharmacological interventions (7, 8). This offers a dynamic readout ideal for stratification and monitoring. Alterations in amino acid metabolism, lipid signaling, energy homeostasis, and oxidative stress markers have been consistently reported in schizophrenia (SCZ) and affective disorders, suggesting that metabolomics features can reveal disease-relevant pathway perturbations (9, 10).
Early metabolomics investigations in psychiatric research were largely exploratory, focusing on identifying candidate metabolites in blood, cerebrospinal fluid (CSF), and urine from small cohorts (11–13). While these studies established proof-of-concept for disease-associated metabolic dysregulation, they often lacked replication, employed heterogeneous analytical methods, and were limited by insufficient sample sizes, hindering clinical translation. More recent efforts have emphasized standardized protocols, targeted validation of previously reported biomarkers, and improved reproducibility. Nonetheless, several challenges remain: pre-analytical and analytical variability, the lack of large-scale longitudinal studies, and the scarce integration of metabolomics data with other omics layers represent key issues.
In this mini-review, we synthesize main findings in psychiatric metabolomics, also providing a critical methodological overview and appraising the field’s major strengths, barriers and priorities for future research. Our overarching goal is to propose a clinically oriented framework that supports progression from biomarker discovery to validation and eventual clinical translation, accelerating the development of metabolomics-informed, personalized psychiatric care.
2 Metabolomics methodology: a brief overview
Metabolomics is defined as the systematic analysis of low-molecular-weight biological compounds (<1.5 kDa) within a specific biological system (6). Thus, the metabolome encompasses the complete set of metabolites – including metabolic intermediates, hormones, signaling molecules, and secondary metabolites – found within a cell, a tissue, a biofluid, or the whole organism (6). These compounds include different chemical classes such as amino acids, lipids, sugars, nucleotides, and organic acids. Unlike the relatively static genome and transcriptome, the metabolome is highly dynamic and time-dependent, changing from second to second in response to gene expression, physiological states, environmental perturbations, and pathophysiological stimuli (14, 15). This dynamic nature puts metabolomics at the functional endpoint of the omics cascade, providing a direct readout of cellular activity and phenotypic expression that integrates upstream genomics, transcriptomics, and proteomics information (16).
A critical aspect of metabolomics research is the choice of the specimen, which varies according to disease and practicality. Blood (whether plasma or serum) and urine represent the most commonly used biofluids in metabolomics due to their non-invasive collection and rich metabolite content reflecting systemic metabolic changes that may be associated with psychiatric conditions. Nonetheless, peripheral metabolites are subject to influence by systemic processes and diet, making them imperfect proxies for brain biochemistry (17). On the other hand, CSF, by virtue of its proximity to central nervous system processes, has been particularly valuable (13), but its collection is invasive and costly, hence less feasible for large cohorts. Saliva represents an emerging biofluid offering completely non-invasive sampling (18), while dried blood spots enable simplified collection, storage, and transport while maintaining metabolite stability (19). The pre-analytical and analytical handling of samples – including fasting state, collection timing, storage conditions, and batch processing – must be rigorously controlled and standardized, as they can significantly affect metabolite profiles (20).
Metabolomics studies employ analytical platforms that detect and quantify these small molecules. The two primary analytical platforms in metabolomics are nuclear magnetic resonance (NMR) spectroscopy and mass spectrometry (MS). Each platform offers complementary strengths and limitations in sensitivity, quantification, and metabolite coverage. NMR detects the most abundant metabolites (typically ≥1 μM) (21) and provides excellent technical reproducibility (22). However, NMR suffers from limited sensitivity and spectral resolution, often resulting in peak overlap that complicates metabolite identification (23). In contrast, MS platforms – including liquid chromatography-MS (LC-MS) and gas chromatography-MS (GC-MS) – detect metabolites that are readily ionizable and offer superior sensitivity, resolution (10³ to 104), and dynamic range compared to NMR (24). Nonetheless, MS platforms are susceptible to technical variability (25). Since many metabolites are uniquely identified by either NMR or MS alone, the combination of NMR and MS in multi-platform approaches significantly improves metabolome coverage (26).
Metabolomics approaches can be broadly divided into two complementary strategies that differ in scope, methodology, and application, i.e., untargeted and targeted analyses (27, 28). Untargeted metabolomics aims to achieve a comprehensive overview of the metabolome by detecting as many metabolites as possible within a biological sample, without prior assumptions about their identity (6). Hence, untargeted metabolomics is suited for hypothesis generation, biomarker discovery, and characterization of global metabolic perturbations (29, 30). However, it faces limitations related to metabolite identification, high-dimensional outputs, and quantitative accuracy, which may hinder reproducibility across studies (31). On the other hand, targeted metabolomics focuses on the selective and quantitative analysis of predefined sets of metabolites (6). Targeted metabolomics offers high sensitivity, specificity, and reproducibility, making it particularly suitable for hypothesis-driven studies and clinical or translational applications (32, 33). Targeted analysis also represents an important part of a metabolomics workflow to validate and expand upon results from untargeted analysis. The main drawback of targeted metabolomics is that only known metabolites with available reference standards can be reliably measured (34). An overview of key methodological aspects in metabolomics research is reported in Table 1.
Table 1. Overview of metabolomics research methodology.
3 Findings from metabolomics studies across main mental disorders: state-of-the-art
Over the last 15 years, an increasing number of both untargeted and targeted metabolomics investigations have been conducted, identifying recurring metabolite alterations implicated in the pathophysiology of mental disorders (Table 2) (7, 9, 35–38). These alterations can often be clustered into some definite biological categories, suggesting that, rather than representing random scattered changes, they converge within specific functional networks to indicate systematic disruptions across key cellular systems.
Table 2. Main classes of metabolites and individual molecules relevant to psychiatric disorders, with their putative role.
In SCZ, metabolomics has revealed consistent patterns of dysregulation involving neurotransmission, oxidation, membrane integrity, and bioenergetics (39–43). Pathway enrichment analyses reported that the most prominently affected processes in SCZ are those related to amino acid processing – including arginine and proline metabolism and alanine/aspartate/glutamate metabolism – which are critical for neurotransmitter synthesis (44). Abnormalities in aromatic amino acid and glycolytic pathways may also account for auditory hallucinations, a core domain of SCZ (45). A compromission of brain protection against oxidative stress is indicated by disrupted glutathione metabolism, with compromised antioxidant defenses and increased vulnerability to oxidative stress (44). Energy-related processes show widespread alterations, including galactose metabolism, glyoxylate/dicarboxylate metabolism, starch/sucrose metabolism, and pantothenate/CoA biosynthesis, suggesting impairment in the brain’s energy production and utilization. Lipid signaling processes – particularly unsaturated fatty acid biosynthesis – are also disrupted, potentially affecting neuronal membrane integrity and, subsequently, downstream signal transduction. Cognitive dysfunction, another core domain of SCZ, has been linked to alterations in neuronal energy metabolism, mitochondrial function, and neurotransmitter cycling (46, 47). Effective antipsychotic treatment seems to modulate many compounds, including several fatty acids, tyrosine, tryptophan, uric acid, lactate, aspartate, glycine, and myo-inositol, indicating partial normalization of brain metabolism as a marker of therapeutic response (48, 49).
As regards bipolar disorder (BD), metabolomic studies have shown widespread alterations in amino acid and lipid metabolisms, energy production, and neurotransmission that underpins both shared and phase-specific pathophysiological mechanisms. Several perturbations in branched-chain (valine, leucine, isoleucine) and aromatic (phenylalanine, tyrosine) amino acid metabolism (50) have been observed. The glycine/serine/threonine pathway seems consistently dysregulated across studies (51), influencing one-carbon metabolism, redox homeostasis, energy production, and cellular signaling. Metabolomics analyses have also confirmed altered tryptophan catabolism in BD (52), with decreased kynurenic acid and increased quinolinic acid in plasma (53). Lipidomics investigations have uncovered decreased plasmalogens and acyl-carnitines alongside elevated triacylglycerols, reflecting mitochondrial dysfunction and altered fatty acid β-oxidation (54). Pathway enrichment analyses have consistently observed disruptions in the tricarboxylic acid (TCA) cycle and glycolysis/gluconeogenesis have been constantly highlighted by blood studies (55) and corroborated by CSF investigations (56), comprehensively pointing towards impaired bioenergetic homeostasis in BD. All these core metabolism dysfunctions are accompanied by phase-specific metabolic signatures that, although stemming from a cross-sectional evidence base, may help to subtype episodes: bipolar depressive episodes show impaired glucose utilization and increased reliance on alternative fuels (β-glucose, glycerol, lactate, acetoacetate, lipids), manic/hypomanic episodes exhibit gut-microbiome and creatine-related shifts, and mixed episodes feature glycine/serine/threonine pathway disruptions (57). Despite the small number of available studies, evidence suggests that metabolomic profiles in subjects with BD seem influenced by psychotropic medications: for instance, lithium treatment resulted in elevated serum levels of L-lactic acid and 3-hydroxymethylglutaric acid compared to second-generation antipsychotic-treated subjects, while reducing linoleic acid and N-acetylglutamic acid concentrations (58).
Metabolomics studies of major depressive disorder (MDD) have consistently revealed perturbations across amino acid, energy-metabolism, lipid, and neurotransmitter pathways. Two-thirds of blood metabolomics studies in MDD converge on decreased levels of many amino acids: tyrosine metabolism and valine, leucine, and isoleucine biosynthesis are constantly among the most perturbed mechanisms, underscoring aromatic and branched-chain amino-acid disruptions (59). The glycine/serine/threonine pathway is also disrupted, implicating one-carbon and redox homeostasis alterations (60). Notable deficits in core bioenergetic processes are observed as well (59): significant perturbations in the TCA cycle (with decreased intermediates pyruvate and fumarate), glyoxylate/dicarboxylate metabolism, and butanoate metabolism indicate widespread energy metabolism dysfunction (61, 62). Consistent changes in phosphatidylethanolamines, lysophosphatidylcholines, ceramides, sphingomyelins, long-chain unsaturated lipids, and pantothenate/CoA biosynthesis also point to membrane alterations, abnormal fatty-acid oxidation, and mitochondrial disfunction (63, 64). Imbalances in excitatory-inhibitory mechanisms, with elevated glutamate and γ-aminobutyric acid (GABA), and reduced glutamine (59), could implicate increased neuronal glutaminase activity or impaired astrocytic uptake, creating a paradox of neurotransmitter excess despite precursor depletion (65). Altered kynurenine pathway (66), with decreased tryptophan, kynurenine and kynurenic acid, along with quinolinic acid, further implicate neuroinflammation and glutamatergic excitotoxicity in MDD, correlating with symptom severity and cognitive deficits (67). Both first-episode drug-naïve MDD (68) and treatment-resistant MDD (69) show peculiar metabolomics features. Antidepressant treatment seems to influence the metabolome, reversing the reduction of brain neurotransmitters caused by depression, modulating inflammatory activation and tryptophan catabolism, and alleviating abnormalities of amino acid, energy, and lipid metabolisms (70), with different metabolites correlating with improved symptoms (71). Notably, ketamine and esketamine modulate several metabolic pathways, including the TCA cycle, glycolysis, the pentose phosphate pathway, lipid metabolism, amino acid metabolism, the kynurenine pathway, and the urea cycle (72).
Metabolomics studies in anxiety disorders underscore perturbations in lipid homeostasis, amino-acid turnover, and TCA-related bioenergetics as promising biomarker candidates. Subjects with anxiety exhibit elevated phosphatidylcholines and ceramides, indicating possible disruption of neuronal membrane integrity, neurotransmitter synthesis, and signal transduction (73). As well, phenylalanine, tyrosine and tryptophan biosynthesis may be related to anxiety disorders (74). Urinary metabolomic profiling also revealed significant alterations in amino acid turnover, tryptophan catabolism and TCA cycle intermediates (73). However, given the high comorbidity between anxiety and depression, metabolomic signatures may partly reflect depressive symptomatology rather than anxiety-specific biology (75). Human studies of the effect of anxiolytic treatments on metabolomics features are lacking.
Metabolic signatures of neuroinflammation, oxidative stress, and bioenergetic deficits have been found in post-traumatic stress disorder (PTSD) among World Trade Center survivors (76) and veterans (77), with evidence of altered glycolytic, TCA, sphingolipid, glutathione, branched-chain amino-acid, and fatty-acid metabolisms. Divergent pathway activation patterns between recent versus chronic PTSD have been found (78). Specific blood metabolites (such as theophylline) have been linked with the risk of developing PTSD (79).
Metabolomic profiling of drug-naïve adults with obsessive-compulsive disorder (OCD) has revealed significant alterations in unsaturated fatty acid, tryptophan metabolism, and glutamate and GABA precursors (80, 81). Notably, plasma docosapentaenoic acid and 5-hydroxytryptophan levels may represent predictors of response to treatment with sertraline (81). These findings are partially substantiated by preclinical evidence (82).
A synthesis of main findings from metabolomics studies in mental disorders is reported in Table 3.
Table 3. Synthesis of findings from metabolomics studies across main mental disorders.
4 Interpretation, strengths, and limitations of metabolomics evidence in psychiatry
A convergent involvement of several pathways across different mental disorders emerges from this body of evidence. Perturbations in branched−chain and aromatic amino acids, together with changes in glutamate, GABA, glycine/serine/threonine and related intermediates, indicate altered neurotransmission (83, 84) across psychosis, mood, anxiety, and compulsivity (85, 86). In parallel, dysregulation of one−carbon and redox−related amino acid pathways may account for gene−expression regulation and vulnerability to oxidative damage observed in psychiatric disorders (87, 88). Consistent abnormalities in glycolysis, TCA cycle intermediates, related anaplerotic routes and markers of high−energy phosphates point to bioenergetics abnormalities (89, 90) that possibly contribute to cognitive symptoms, anergia, and diminished resilience. Alterations in acylcarnitines and other fatty acid-related metabolites additionally suggest that the capacity to flexibly switch between energy sources may be impaired in at least a subset of patients (91). Lipidomic findings are in line with disruptions of membrane homeostasis and lipid−mediated signaling (92). Finally, consistent involvement of tryptophan catabolism supports increased neurotoxicity (93).
Overall, metabolomics studies suggest that such biochemical alterations – largely shared across disorders – likely reflect both a genuine transdiagnostic pathophysiology, with convergent metabolic perturbations aligning with dimensional constructs that transcend diagnostic boundaries, and nosological shortcomings, with high diagnostic comorbidity and symptom overlap not cleanly separating distinct pathophysiologies. Also, most studies focus on SCZ, BD, and MDD, and findings are relatively well-replicated in multiple independent cohorts, providing an evidence base that supports biological validity and potential clinical utility. On the other hand, evidence for anxiety disorders, OCD, and PTSD remains preliminary, typically based on no more than three primary studies with limited cross-study replication. Sample sizes are small and often insufficient: while pilot discovery screens may require only 20–30 participants per group, validation studies demand substantially larger cohorts (~300 subjects for 0.95 power in multivariate settings), yet most investigations fall short of these thresholds (94, 95). First-episode, medication-naïve samples and stratification by drug class should be prioritized. The predominance of cross‐sectional designs further limits the differentiation of trait biomarkers from state‐dependent or treatment‐induced metabolic changes, not allowing a thorough characterization of temporal dynamics.
Lack of independent replication and external validation undermines the generalizability of reported findings. The reliance on peripheral biofluids introduces uncertainty about the correspondence between systemic and central biochemistry. Residual confounding by lifestyle factors, adiposity, smoking and medication can all generate apparent similarity in metabolite profiles. In particular, treatment-response studies are scarce and largely observational (96), limiting causal inference about whether metabolic shifts reflect therapeutic mechanisms or drug confounding. Pre‐analytical variables (biofluid type, fasting status and sampling/storage/processing protocols) vary substantially (97), as do analytical platforms: NMR is limited by lower sensitivity and resolution (23), while MS can suffer from variability that affects reproducibility (31). High-dimensional metabolomics data, with the number of variables greatly exceeding the number of samples [the “curse of dimensionality” (98)], require complex pre-processing and pose the risk of model overfitting. This requires robust computational frameworks (99): advancing data analysis, for instance implementing advanced ML approaches, can enhance phenotype discrimination and produce more refined predictive models (100). Explainable AI models (101, 102) and open data sharing (103) can improve interpretability and accelerate validation efforts.
Therefore, while metabolomics offers unprecedented insights into the molecular underpinnings of mental disorders by providing objective, high-throughput profiling of small-molecule metabolites, laying the groundwork for precision psychiatry, this literature is best interpreted as promising but preliminary.
Although there is recurrent evidence for small, biologically anchored panels (e.g., those involving amino acids and kynurenine metabolites), only a few metabolomic signatures in psychiatry can currently be considered “near translational”, and none is yet ready for routine clinical use. Candidate metabolites/panels still face major barriers: untargeted workflows are costly and technically demanding, targeted assays are not yet standardized across laboratories, and no study has conclusively shown that adding metabolomics testing to existing clinical assessments improves outcomes in a cost-effective way. Combining complementary techniques can overcome the limits of single analytical platforms, yielding more comprehensive biomarker panels (26, 50). Moreover, multi-omics approaches, jointly analyzing metabolomics with genomics, transcriptomics, and proteomics, can generate robust composite signatures that surpass single-omics to yield deeper mechanistic insights (99, 104), and integrating metabolomics with genetics, neuroimaging and digital data can provide multimodal combinations that are more likely to yield clinically useful stratification and prediction than any single modality alone (105, 106).
In general, study designs should be tailored to clinical utility, focusing on prospective, outcome-oriented biomarker validation to ensure that findings are directly translatable into meaningful clinical interventions, for instance identifying metabolomics predictors of treatment response (107).
5 Conclusions
Metabolomics offers unprecedented insights into the molecular underpinnings of mental disorders by providing objective, high-throughput profiling of small-molecule metabolites, laying the groundwork for precision psychiatry. The high dimensionality of metabolomics data has the potential to be used to identify metabolic signatures that distinguish diagnostic groups and predict symptom trajectories, especially when considered not as isolated markers but within panels. In this direction, composite metabolite scores and pathway enrichment further help gain insights into metabolic abnormalities (44). Beyond diagnostics, metabolomics monitoring can offer a dynamic tool for individualized treatment optimization by tracking biochemical responses to pharmacotherapy (96). Comprehensively, metabolomics promises to deliver tailored diagnostic panels, adaptive monitoring strategies, and mechanism-driven therapies that could move psychiatry beyond trial-and-error prescribing towards truly personalized mental health care.
Metabolomics has matured from small, proof-of-principle studies to large multi-platform investigations that consistently implicate several biochemical domains across psychiatric disorders. To date, studies have identified alterations in amino acid, lipid, energy, and inflammatory pathways in severe mental disorders. Yet, despite the encouraging evidence, the field faces substantial hurdles, with methodological challenges currently limiting harmonization and clinical translation of these findings. When these components are in place, metabolomics can become a key pillar of biologically informed, personalized psychiatric care, improving diagnosis, subtype stratification, treatment, and clinical monitoring.
Author contributions
DC: Data curation, Methodology, Conceptualization, Investigation, Writing – original draft. CB: Writing – review & editing, Data curation, Conceptualization, Investigation. GCu: Investigation, Data curation, Conceptualization, Writing – review & editing. PDF: Conceptualization, Writing – review & editing, Project administration, Funding acquisition. RdF: Data curation, Writing – review & editing, Conceptualization, Investigation. UA: Writing – review & editing, Conceptualization, Funding acquisition, Project administration. LP: Writing – review & editing, Conceptualization, Investigation, Data curation. GCa: Methodology, Conceptualization, Supervision, Writing – review & editing. FB: Project administration, Methodology, Conceptualization, Funding acquisition, Writing – review & editing.
Funding
The author(s) declared financial support was received for this work and/or its publication. This study is funded by the Italian Ministry of University and Research (MUR) as a Research Projects of Significant National Interest (PRIN – Progetti di Rilevante Interesse Nazionale) – 2022 call – Prot. 2022C7AL7F.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author(s) declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.
Generative AI statement
The author(s) declared that generative AI was used in the creation of this manuscript. ChatGPT (OpenAI, San Francisco, CA, USA) was used to refine the language and structure of the manuscript. The authors reviewed and take full responsibility for the content of this publication.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1. Uher R and Zwicker A. Etiology in psychiatry: embracing the reality of poly-gene-environmental causation of mental illness. World Psychiatry. (2017) 16:121–29. doi: 10.1002/wps.20436
2. Stein DJ, Shoptaw SJ, Vigo DV, Lund C, Cuijpers P, Bantjes J, et al. Psychiatric diagnosis and treatment in the 21st century: paradigm shifts versus incremental integration. World Psychiatry. (2022) 21:393–414. doi: 10.1002/wps.20998
3. García-Gutiérrez MS, Navarrete F, Sala F, Gasparyan A, Austrich-Olivares A, and Manzanares J. Biomarkers in psychiatry: concept, definition, types and relevance to the clinical reality. Front Psychiatry. (2020) 11:432. doi: 10.3389/fpsyt.2020.00432
4. Allsopp K, Read J, Corcoran R, and Kinderman P. Heterogeneity in psychiatric diagnostic classification. Psychiatry Res. (2019) 279:15–22. doi: 10.1016/j.psychres.2019.07.005
5. Kas MJH, Penninx BWJH, Knudsen GM, Cuthbert B, Falkai P, Sachs GS, et al. Precision psychiatry roadmap: towards a biology-informed framework for mental disorders. Mol Psychiatry. (2025) 30:3846–55. doi: 10.1038/s41380-025-03070-5
6. Johnson C, Ivanisevic J, and Siuzdak G. Metabolomics: beyond biomarkers and towards mechanisms. Nat Rev Mol Cell Biol. (2016) 17:451–59. doi: 10.1038/nrm.2016.25
7. Shih PB. Metabolomics biomarkers for precision psychiatry. Adv Exp Med Biol. (2019) 1161:101–13. doi: 10.1007/978-3-030-21735-8_10
8. Patti G, Yanes O, and Siuzdak G. Metabolomics: the apogee of the omics trilogy. Nat Rev Mol Cell Biol. (2012) 13:263–69. doi: 10.1038/nrm3314
9. Konjevod M, Sáiz J, Bordoy L, Strac DS, Taha AY, Lanceros-Méndez S, et al. Validated metabolomic biomarkers in psychiatric disorders: a narrative review. Mol Med. (2025) 31:254. doi: 10.1186/s10020-025-01258-7
10. Pedrini M, Cao B, Nani JVS, Cerqueira RO, Mansur RB, Tasic L, et al. Advances and challenges in development of precision psychiatry through clinical metabolomics on mood and psychotic disorders. Prog Neuropsychopharmacol Biol Psychiatry. (2019) 93:182–88. doi: 10.1016/j.pnpbp.2019.03.010
11. Zheng P, Wang Y, Chen L, Yang D, Meng H, Zhou D, et al. Identification and validation of urinary metabolite biomarkers for major depressive disorder. Mol Cell Proteomics. (2013) 12:207–14. doi: 10.1074/mcp.M112.021816
12. He Y, Yu Z, Giegling I, Xie L, Hartmann AM, Prehn C, et al. Schizophrenia shows a unique metabolomics signature in plasma. Transl Psychiatry. (2012) 2:e149. doi: 10.1038/tp.2012.76
13. Holmes E, Tsang TM, Huang JT, Leweke FM, Koethe D, Gerth CW, et al. Metabolic profiling of CSF: evidence that early intervention may impact on disease progression and outcome in schizophrenia. PloS Med. (2006) 3:e327. doi: 10.1371/journal.pmed.0030327
14. Agueusop I, Musholt PB, Klaus B, Hightower K, and Kannt A. Short-term variability of the human serum metabolome depending on nutritional and metabolic health status. Sci Rep. (2020) 10:16310. doi: 10.1038/s41598-020-72914-7
15. Link H, Fuhrer T, Gerosa L, Zamboni N, and Sauer U. Real-time metabolome profiling of the metabolic switch between starvation and growth. Nat Methods. (2015) 12:1091–97. doi: 10.1038/nmeth.3584
16. Nava AA and Arboleda VA. The omics era: a nexus of untapped potential for Mendelian chromatinopathies. Hum Genet. (2024) 143:475–95. doi: 10.1007/s00439-023-02560-2
17. Yin B, Cai Y, Teng T, Wang X, Liu X, Li X, et al. Identifying plasma metabolic characteristics of major depressive disorder, bipolar disorder, and schizophrenia in adolescents. Transl Psychiatry. (2024) 14:163. doi: 10.1038/s41398-024-02886-z
18. Gardner A, Carpenter G, and So PW. Salivary metabolomics: from diagnostic biomarker discovery to investigating biological function. Metabolites. (2020) 10:47. doi: 10.3390/metabo10020047
19. Cui HN, Shi F, Huang G, He Y, Yu S, Liu L, et al. Evaluation of metabolite stability in dried blood spot stored at different temperatures and times. Sci Rep. (2024) 14:30964. doi: 10.1038/s41598-024-82041-2
20. Stevens VL, Hoover E, Wang Y, and Zanetti KA. Pre-analytical factors that affect metabolite stability in human urine, plasma, and serum: A review. Metabolites. (2019) 9:156. doi: 10.3390/metabo9080156
21. Wishart DS, Cheng LL, Copié V, Edison AS, Eghbalnia HR, Hoch JC, et al. NMR and metabolomics–A roadmap for the future. Metabolites. (2022) 12:678. doi: 10.3390/metabo12080678
22. Li-Gao R, Hughes DA, le Cessie S, de Mutsert R, den Heijer M, Rosendaal FR, et al. Assessment of reproducibility and biological variability of fasting and postprandial plasma metabolite concentrations using 1H NMR spectroscopy. PloS One. (2019) 14:e0218549. doi: 10.1371/journal.pone.0218549
23. Emwas AH, Roy R, McKay RT, Tenori L, Saccenti E, Gowda GAN, et al. NMR spectroscopy for metabolomics research. Metabolites. (2019) 9:123. doi: 10.3390/metabo9070123
24. Hajnajafi K and Iqbal MA. Mass-spectrometry based metabolomics: an overview of workflows, strategies, data analysis and applications. Proteome Sci. (2025) 23:5. doi: 10.1186/s12953-025-00241-8
25. Thomas SN, French D, Jannetto PJ, Rappold BA, and Clarke WA. Liquid chromatography-tandem mass spectrometry for clinical diagnostics. Nat Rev Methods Primers. (2022) 2:96. doi: 10.1038/s43586-022-00175-x
26. Bhinderwala F, Wase N, DiRusso C, and Powers R. Combining mass spectrometry and NMR improves metabolite detection and annotation. J Proteome Res. (2018) 17:4017–22. doi: 10.1021/acs.jproteome.8b00567
27. Nedic Erjavec G, Konjevod M, Nikolac Perković M, Svob Strac D, Tudor L, Barbas C, et al. Short overview on metabolomic approach and redox changes in psychiatric disorders. Redox Biol. (2018) 14:178–86. doi: 10.1016/j.redox.2017.09.002
28. Ribbenstedt A, Ziarrusta H, and Benskin JP. Development, characterization and comparisons of targeted and non-targeted metabolomics methods. PloS One. (2018) 13:e0207082. doi: 10.1371/journal.pone.0207082
29. Wurth R, Turgeon C, Stander Z, and Oglesbee D. An evaluation of untargeted metabolomics methods to characterize inborn errors of metabolism. Mol Genet Metab. (2024) 141:108–15. doi: 10.1016/j.ymgme.2023.108115
30. González-Domínguez R, González-Domínguez Á, Segundo C, Schwarz M, Sayago A, Mateos RM, et al. High-throughput metabolomics based on direct mass spectrometry analysis in biomedical research. Methods Mol Biol. (2019) 1978:27–38. doi: 10.1007/978-1-4939-9236-2_3
31. Lin Y, Caldwell GW, Li Y, Lang W, and Masucci J. Inter-laboratory reproducibility of an untargeted metabolomics GC-MS assay for analysis of human plasma. Sci Rep. (2020) 10:10918. doi: 10.1038/s41598-020-67939-x
32. Hu Z, Shen F, Liu Y, Zhong Z, Chen Y, Xia Z, et al. Targeted metabolomics reveals novel diagnostic biomarkers for colorectal cancer. Mol Oncol. (2025) 19:1737–50. doi: 10.1002/1878-0261.13791
33. Siskos AP, Jain P, Römisch-Margl W, Bennett M, Achaintre D, Asad Y, et al. Interlaboratory reproducibility of a targeted metabolomics platform for analysis of human serum and plasma. Anal Chem. (2017) 89:656–65. doi: 10.1021/acs.analchem.6b02930
34. Ghafari N and Sleno L. Challenges and recent advances in quantitative mass spectrometry-based metabolomics. Anal Sci Adv. (2024) 5:e2400007. doi: 10.1002/ansa.202400007
35. Jaber M, Kahwaji H, Nasr S, Baz R, Kim YK, and Fakhoury M. Precision medicine in depression: the role of proteomics and metabolomics in personalized treatment approaches. Adv Exp Med Biol. (2024) 1456:359–78. doi: 10.1007/978-981-97-4402-2_18
36. Verma A, Inslicht SS, and Bhargava A. Gut-brain axis: role of microbiome, metabolomics, hormones, and stress in mental health disorders. Cells. (2024) 13:1436. doi: 10.3390/cells13171436
37. Inam ME, Fernandes BS, Salagre E, Grande I, Vieta E, Quevedo J, et al. The kynurenine pathway in major depressive disorder, bipolar disorder, and schizophrenia: a systematic review and meta-analysis of cerebrospinal fluid studies. Braz J Psychiatry. (2023) 45:343–55. doi: 10.47626/1516-4446-2022-2973
38. Guerreiro Costa LNF, Carneiro BA, Alves GS, Lins Silva DH, Faria Guimaraes D, Souza LS, et al. Metabolomics of major depressive disorder: A systematic review of clinical studies. Cureus. (2022) 14:e23009. doi: 10.7759/cureus.23009
39. Wang X, Xie J, Ma H, Li G, Li M, Li S, et al. The relationship between alterations in plasma metabolites and treatment responses in antipsychotic-naïve female patients with schizophrenia. World J Biol Psychiatry. (2024) 25:106–15. doi: 10.1080/15622975.2023.2271965
40. Liu Y, Song X, Liu X, Pu J, Gui S, Xu S, et al. Alteration of lipids and amino acids in plasma distinguish schizophrenia patients from controls: a targeted metabolomics study. Psychiatry Clin Neurosci. (2021) 75:138–44. doi: 10.1111/pcn.13194
41. Okamoto N, Ikenouchi A, Watanabe K, Igata R, Fujii R, and Yoshimura R. A metabolomics study of serum in hospitalized patients with chronic schizophrenia. Front Psychiatry. (2021) 12:763547. doi: 10.3389/fpsyt.2021.763547
42. Cui G, Qing Y, Li M, Sun L, Zhang J, Feng L, et al. Salivary metabolomics reveals that metabolic alterations precede the onset of schizophrenia. J Proteome Res. (2021) 20:5010–23. doi: 10.1021/acs.jproteome.1c00504
43. Koike S, Bundo M, Iwamoto K, Suga M, Kuwabara H, Ohashi Y, et al. A snapshot of plasma metabolites in first-episode schizophrenia: a capillary electrophoresis time-of-flight mass spectrometry study. Transl Psychiatry. (2014) 4:e379. doi: 10.1038/tp.2014.19
44. Yao G, Zeng J, Huang Y, Lu H, Ping J, Wan J, et al. Discovery of biological markers for schizophrenia based on metabolomics: a systematic review. Front Psychiatry. (2025) 16:1540260. doi: 10.3389/fpsyt.2025.1540260
45. Li X, Yang C, Liang X, Li D, Zhou Z, Xiao H, et al. Metabolomics and cytokine analysis for identification of schizophrenia with auditory hallucination. Clin Invest Med. (2022) 45:E39–48. doi: 10.25011/cim.v45i2.38096
46. Liu H, Huang Z, Zhang X, He Y, Gu S, Mo D, et al. Association between lipid metabolism and cognitive function in patients with schizophrenia. Front Psychiatry. (2022) 13:1013698. doi: 10.3389/fpsyt.2022.1013698
47. Jiang Y, Sun X, Hu M, Zhang L, Zhao N, Shen Y, et al. Plasma metabolomics of schizophrenia with cognitive impairment: a pilot study. Front Psychiatry. (2022) 13:950602. doi: 10.3389/fpsyt.2022.950602
48. Xuan J, Pan G, Qiu Y, Yang L, Su M, Liu Y, et al. Metabolomic profiling to identify potential serum biomarkers for schizophrenia and risperidone action. J Proteome Res. (2011) 10:5433–43. doi: 10.1021/pr2006796
49. Song M, Liu Y, Zhou J, Shi H, Su X, Shao M, et al. Potential plasma biomarker panels identification for the diagnosis of first-episode schizophrenia and monitoring antipsychotic monotherapy with the use of metabolomics analyses. Psychiatry Res. (2023) 321:115070. doi: 10.1016/j.psychres.2023.115070
50. Chen JJ, Liu Z, Fan SH, Yang DY, Zheng P, Shao WH, et al. Combined application of NMR- and GC-MS-based metabonomics yields a superior urinary biomarker panel for bipolar disorder. Sci Rep. (2014) 4:5855. doi: 10.1038/srep05855
51. Chen J, Amdanee N, Zuo X, Wang Y, Gong M, Yang Y, et al. Biomarkers of bipolar disorder based on metabolomics: A systematic review. J Affect Disord. (2024) 350:492–503. doi: 10.1016/j.jad.2024.01.033
52. Bartoli F, Cioni RM, Cavaleri D, Callovini T, Crocamo C, Misiak B, et al. The association of kynurenine pathway metabolites with symptom severity and clinical features of bipolar disorder: An overview. Eur Psychiatry. (2022) 65:e82. doi: 10.1192/j.eurpsy.2022.2340
53. Nakayama T, Umehara H, Mawatari K, Tomioka Y, Yoshida T, Matsuda H, et al. Alterations of blood plasma metabolites, including kynurenine and tryptophan, in bipolar disorder. Neuropsychiatr Dis Treat. (2025) 21:1067–73. doi: 10.2147/NDT.S508021
54. Tomasik J, Harrison SJ, Rustogi N, Olmert T, Barton-Owen G, Han SYS, et al. Metabolomic biomarker signatures for bipolar and unipolar depression. JAMA Psychiatry. (2024) 81:101–06. doi: 10.1001/jamapsychiatry.2023.4096
55. Cao T, Xu B, Li S, Qiu Y, Chen JD, Wu HS, et al. Bioenergetic biomarkers as predictive indicators and their relationship with cognitive function in newly diagnosed, drug-naïve patients with bipolar disorder. Transl Psychiatry. (2025) 15:148. doi: 10.1038/s41398-025-03367-7
56. Smedler E, Salehi AM, Pelanis A, Andreazza A, Pålsson E, Sparding T, et al. Metabolomics analysis of cerebrospinal fluid suggests citric acid cycle aberrations in bipolar disorder. Neurosci Appl. (2022) 1:100108. doi: 10.1016/j.nsa.2022.100108
57. Guo Q, Jia J, Sun XL, Yang H, and Ren Y. Comparing the metabolic pathways of different clinical phases of bipolar disorder through metabolomics studies. Front Psychiatry. (2024) 14:1319870. doi: 10.3389/fpsyt.2023.1319870
58. Burghardt KJ, Evans SJ, Wiese KM, and Ellingrod VL. An untargeted metabolomics analysis of antipsychotic use in bipolar disorder. Clin Transl Sci. (2015) 8:432–40. doi: 10.1111/cts.12324
59. Wang Y, Cai X, Ma Y, Yang Y, Pan CW, Zhu X, et al. Metabolomics on depression: A comparison of clinical and animal research. J Affect Disord. (2024) 349:559–68. doi: 10.1016/j.jad.2024.01.053
60. Dong T, Wang X, Jia Z, Yang J, and Liu Y. Assessing the associations of 1,400 blood metabolites with major depressive disorder: A Mendelian randomization study. Front Psychiatry. (2024) 15:1391535. doi: 10.3389/fpsyt.2024.1391535
61. Hung CI, Lin G, Chiang MH, and Chiu CY. Metabolomics-based discrimination of patients with remitted depression from healthy controls using 1H-NMR spectroscopy. Sci Rep. (2021) 11:15608. doi: 10.1038/s41598-021-95221-1
62. Drevets WC, Price JL, and Furey ML. Brain structural and functional abnormalities in mood disorders: implications for neurocircuitry models of depression. Brain Struct Funct. (2008) 213:93–118. doi: 10.1007/s00429-008-0189-x
63. Liu M, Ma W, He Y, Sun Z, and Yang J. Recent progress in mass spectrometry-based metabolomics in major depressive disorder research. Molecules. (2023) 28:7430. doi: 10.3390/molecules28217430
64. Wu Z, Yu H, Tian Y, Wang Y, He Y, Lan T, et al. Non-targeted metabolomics profiling of plasma samples from patients with major depressive disorder. Front Psychiatry. (2022) 12:810302. doi: 10.3389/fpsyt.2021.810302
65. Abdallah CG, Jiang L, De Feyter HM, Fasula M, Krystal JH, Rothman DL, et al. Glutamate metabolism in major depressive disorder. Am J Psychiatry. (2014) 171:1320–27. doi: 10.1176/appi.ajp.2014.14010067
66. Bartoli F, Cioni RM, Callovini T, Cavaleri D, Crocamo C, and Carrà G. The kynurenine pathway in schizophrenia and other mental disorders: Insight from meta-analyses on the peripheral blood levels of tryptophan and related metabolites. Schizophr Res. (2021) 232:61–2. doi: 10.1016/j.schres.2021.04.008
67. Lee S, Lee YR, Mun S, Yun Y, Kang HG, and Lee J. Biomarker for diagnosis and monitoring of treatment response in major depressive disorder: changes in serum L-glutamine levels. BioMed Chromatogr. (2025) 39:e70197. doi: 10.1002/bmc.70197
68. Gan X, Li X, Cai Y, Yin B, Pan Q, Teng T, et al. Metabolic features of adolescent major depressive disorder: A comparative study between treatment-resistant depression and first-episode drug-naïve depression. Psychoneuroendocrinology. (2024) 167:107086. doi: 10.1016/j.psyneuen.2024
69. Pan LA, Naviaux JC, Wang L, Li K, Monk JM, Lingampelly SS, et al. Metabolic features of treatment-refractory major depressive disorder with suicidal ideation. Transl Psychiatry. (2023) 13:393. doi: 10.1038/s41398-023-02696-9
70. Pu J, Liu Y, Gui S, Tian L, Yu Y, Wang D, et al. Effects of pharmacological treatment on metabolomic alterations in animal models of depression. Transl Psychiatry. (2022) 12:175. doi: 10.1038/s41398-022-01947-5
71. Bhattacharyya S, MahmoudianDehkordi S, Sniatynski MJ, Belenky M, Marur VR, Rush AJ, et al. Metabolomics signatures of serotonin reuptake inhibitor (escitalopram), serotonin norepinephrine reuptake inhibitor (duloxetine) and cognitive-behavioral therapy on key neurotransmitter pathways in major depressive disorder. J Affect Disord. (2025) 375:397–405. doi: 10.1016/j.jad.2025.01.064
72. Cavaleri D, Riboldi I, Crocamo C, Paglia G, Carrà G, and Bartoli F. Evidence from preclinical and clinical metabolomics studies on the antidepressant effects of ketamine and esketamine. Neurosci Lett. (2024) 831:137791. doi: 10.1016/j.neulet.2024.137791
73. Humer E, Pieh C, and Probst T. Metabolomic biomarkers in anxiety disorders. Int J Mol Sci. (2020) 21:4784. doi: 10.3390/ijms21134784
74. Kui H, Su H, Wang Q, Liu C, Li Y, Tian Y, et al. Serum metabolomics study of anxiety disorder patients based on LC-MS. Clin Chim Acta. (2022) 533:131–43. doi: 10.1016/j.cca.2022.06.022
75. de Kluiver H, Jansen R, Milaneschi Y, Bot M, Giltay EJ, Schoevers R, et al. Metabolomic profiles discriminating anxiety from depression. Acta Psychiatr Scand. (2021) 144:178–93. doi: 10.1111/acps.13310
76. Kuan PF, Yang X, Kotov R, Clouston S, Bromet E, and Luft BJ. Metabolomics analysis of post-traumatic stress disorder symptoms in World Trade Center responders. Transl Psychiatry. (2022) 12:174. doi: 10.1038/s41398-022-01940-y
77. Mellon SH, Bersani FS, Lindqvist D, Hammamieh R, Donohue D, Dean K, et al. Metabolomic analysis of male combat veterans with post traumatic stress disorder. PloS One. (2019) 14:e0213839. doi: 10.1371/journal.pone.0213839
78. Muhie S, Gautam A, Misganaw B, Yang R, Mellon SH, Hoke A, et al. Integrated analysis of proteomics, epigenomics and metabolomics data revealed divergent pathway activation patterns in the recent versus chronic post-traumatic stress disorder. Brain Behav Immun. (2023) 113:303–16. doi: 10.1016/j.bbi.2023.07.015
79. Wei Y, Huang L, Sui J, Liu C, and Qi M. Human blood metabolites and risk of post-traumatic stress disorder: A Mendelian randomization study. J Affect Disord. (2025) 372:227–33. doi: 10.1016/j.jad.2024.12.029
80. Chen G, Zhao X, Xie M, Chen H, Shao C, Zhang X, et al. Serum metabolites and inflammation predict brain functional connectivity changes in Obsessive-Compulsive disorder. Brain Behav Immun. (2025) 126:113–25. doi: 10.1016/j.bbi.2025.01.013
81. Li Z, Gao J, Lin L, Zheng Z, Yan S, Wang W, et al. Untargeted metabolomics analysis in drug-naïve patients with severe obsessive-compulsive disorder. Front Neurosci. (2023) 17:1148971. doi: 10.3389/fnins.2023.1148971
82. Abreu AC, Mora S, Tristán AI, Martín-González E, Prados-Pardo Á, Moreno M, et al. NMR-based metabolomics and fatty acid profiles to unravel biomarkers in preclinical animal models of compulsive behavior. J Proteome Res. (2022) 21:612–22. doi: 10.1021/acs.jproteome.1c00857
83. Goulty M, Botton-Amiot G, Rosato E, Sprecher SG, and Feuda R. The monoaminergic system is a bilaterian innovation. Nat Commun. (2023) 14:3284. doi: 10.1038/s41467-023-39030-2
84. Yudkoff M. Interactions in the metabolism of glutamate and the branched-chain amino acids and ketoacids in the CNS. Neurochem Res. (2017) 42:10–8. doi: 10.1007/s11064-016-2057-z
85. Forsyth JK and Lewis DA. Mapping the consequences of impaired synaptic plasticity in schizophrenia through development: an integrative model for diverse clinical features. Trends Cognit Sci. (2017) 21:760–78. doi: 10.1016/j.tics.2017.06.006
86. Duman RS, Aghajanian GK, Sanacora G, and Krystal JH. Synaptic plasticity and depression: new insights from stress and rapid-acting antidepressants. Nat Med. (2016) 22:238–49. doi: 10.1038/nm.4050
87. Bou Ghanem A, Hussayni Y, Kadbey R, Ratel Y, Yehya S, Khouzami L, et al. Exploring the complexities of 1C metabolism: implications in aging and neurodegenerative diseases. Front Aging Neurosci. (2024) 15:1322419. doi: 10.3389/fnagi.2023.1322419
88. Wang W, Wu Z, Dai Z, Yang Y, Wang J, and Wu G. Glycine metabolism in animals and humans: implications for nutrition and health. Amino Acids. (2013) 45:463–77. doi: 10.1007/s00726-013-1493-1
89. Harris JJ, Jolivet R, and Attwell D. Synaptic energy use and supply. Neuron. (2012) 75:762–77. doi: 10.1016/j.neuron.2012.08.019
90. Kann O and Kovács R. Mitochondria and neuronal activity. Am J Physiol Cell Physiol. (2007) 292:C641–57. doi: 10.1152/ajpcell.00222.2006
91. Dambrova M, Makrecka-Kuka M, Kuka J, Vilskersts R, Nordberg D, Attwood MM, et al. Acylcarnitines: nomenclature, biomarkers, therapeutic potential, drug targets, and clinical trials. Pharmacol Rev. (2022) 74:506–51. doi: 10.1124/pharmrev.121.000408
92. Merrill AH Jr. Sphingolipid and glycosphingolipid metabolic pathways in the era of sphingolipidomics. Chem Rev. (2011) 111:6387–422. doi: 10.1021/cr2002917
93. Savitz J. The kynurenine pathway: a finger in every pie. Mol Psychiatry. (2020) 25:131–47. doi: 10.1038/s41380-019-0414-4
94. Tolstikov V, Moser AJ, Sarangarajan R, Narain NR, and Kiebish MA. Current status of metabolomic biomarker discovery: impact of study design and demographic characteristics. Metabolites. (2020) 10:224. doi: 10.3390/metabo10060224
95. Dunn WB, Lin W, Broadhurst D, Begley P, Brown M, Zelená E, et al. Molecular phenotyping of a UK population: defining the human serum metabolome. Metabolomics. (2015) 11:9–26. doi: 10.1007/s11306-014-0707-1
96. Caspani G, Turecki G, Lam RW, Milev RV, Frey BN, MacQueen GM, et al. Metabolomic signatures associated with depression and predictors of antidepressant response in humans: A CAN-BIND-1 report. Commun Biol. (2021) 4:903. doi: 10.1038/s42003-021-02421-6
97. Thachil A, Wang L, Mandal R, Wishart D, and Blydt-Hansen T. An overview of pre-analytical factors impacting metabolomics analyses of blood samples. Metabolites. (2024) 14:474. doi: 10.3390/metabo14090474
98. Skaf Y and Laubenbacher R. Topological data analysis in biomedicine: A review. J BioMed Inform. (2022) 130:104082. doi: 10.1016/j.jbi.2022.104082
99. Argelaguet R, Cuomo ASE, Stegle O, and Marioni JC. Computational principles and challenges in single-cell data integration. Nat Biotechnol. (2021) 39:1202–15. doi: 10.1038/s41587-021-00895-7
100. Joyce JB, Grant CW, Liu D, MahmoudianDehkordi S, Kaddurah-Daouk R, Skime M, et al. Multi-omics driven predictions of response to acute phase combination antidepressant therapy: a machine learning approach with cross-trial replication. Transl Psychiatry. (2021) 11:513. doi: 10.1038/s41398-021-01632-z
101. Agamah FE, Bayjanov JR, Niehues A, Njoku KF, Skelton M, Mazandu GK, et al. Computational approaches for network-based integrative multi-omics analysis. Front Mol Biosci. (2022) 9:967205. doi: 10.3389/fmolb.2022.967205
102. Galal A, Talal M, and Moustafa A. Applications of machine learning in metabolomics: Disease modeling and classification. Front Genet. (2022) 13:1017340. doi: 10.3389/fgene.2022.1017340
103. Yurekten O, Payne T, Tejera N, Amaladoss FX, Martin C, Williams M, et al. MetaboLights: open data repository for metabolomics. Nucleic Acids Res. (2024) 52:D640–46. doi: 10.1093/nar/gkad1045
104. Heo YJ, Hwa C, Lee GH, Park JM, and An JY. Integrative multi-omics approaches in cancer research: from biological networks to clinical subtypes. Mol Cells. (2021) 44:433–43. doi: 10.14348/molcells.2021.0042
105. Brzenczek C, Klopfenstein Q, Hähnel T, Fröhlich H, Glaab E, and NCER-PD Consortium. Integrating digital gait data with metabolomics and clinical data to predict outcomes in Parkinson’s disease. NPJ Digit Med. (2024) 7:235. doi: 10.1038/s41746-024-01236-z
106. Oudin A, Maatoug R, Bourla A, Ferreri F, Bonnot O, Millet B, et al. Digital phenotyping: data-driven psychiatry to redefine mental health. J Med Internet Res. (2023) 25:e44502. doi: 10.2196/44502
107. Bartoli F, Cavaleri D, Riboldi I, Crocamo C, de Filippis R, Zandonella Callegher R, et al. The resistant depression response to esketamine assessing metabolomics (ReDREAM) project–untargeted metabolomics to identify biomarkers of treatment response to intranasal esketamine in individuals with treatment-resistant depression: A study protocol. Alpha Psychiatry. (2024) 25:555–60. doi: 10.5152/alphapsychiatry.2024.241549
Keywords: anxiety disorders, biomarkers, bipolar disorder, major depressive disorder, metabolomics, obsessive-compulsive disorder, post-traumatic stress disorder, schizophrenia
Citation: Cavaleri D, Bassetti C, Cucchi G, De Fazio P, de Filippis R, Albert U, Pellegrini L, Carrà G and Bartoli F (2026) Metabolomics biomarkers for precision psychiatry. Front. Psychiatry 17:1736206. doi: 10.3389/fpsyt.2026.1736206
Received: 30 October 2025; Accepted: 30 January 2026; Revised: 26 January 2026;
Published: 13 February 2026.
Edited by:
Andrea Fiorillo, University of Campania Luigi Vanvitelli, ItalyReviewed by:
Marwa Zafarullah, Stanford University, United StatesSalvatore Cipolla, University of Campania ‘Luigi Vanvitelli’, Italy
Copyright © 2026 Cavaleri, Bassetti, Cucchi, De Fazio, de Filippis, Albert, Pellegrini, Carrà and Bartoli. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Daniele Cavaleri, ZC5jYXZhbGVyaTFAY2FtcHVzLnVuaW1pYi5pdA==