Validated metabolomic biomarkers in psychiatric disorders: a narrative review

Abstract

Schizophrenia, major depressive disorder, bipolar disorder and posttraumatic stress disorder are severe and profoundly debilitating mental illnesses. Due to their heterogeneity and polygenic nature, the metabolic pathways and biological mechanisms underlying these conditions remain elusive. Consequently, diagnosing psychiatric disorders is a complex and multifaceted process, relying on clinical assessment and standardized diagnostic criteria. There is a growing demand to identify and integrate potential biomarkers for these disorders, especially for early diagnosis, to enhance diagnostic accuracy and complement existing diagnostic tools. Validating potential diagnostic biomarkers is essential to ensure their accuracy, reliability, generalizability, and clinical utility.In this article we provide a comprehensive review of validated metabolomics research, focusing on both the specific psychiatric conditions and the types of validation performed. Our scope is limited to peer-reviewed studies that include studies that performed validation studies in independent cohorts, cross-validation, or external validation. Due to the lack of published research, most of these validation studies have concentrated on major depressive disorder and schizophrenia, with fewer studies addressing bipolar disorder and posttraumatic stress disorder.Biomarkers are considered as validated if they demonstrated reproducibility in additional cohorts and biological relevance across independent datasets. However, several limitations must be acknowledged, including the heterogeneity in study design, small sample sizes, different analytical platforms, and inconsistent validation criteria across studies. Published results reveal that potential metabolomics biomarkers pertain to diverse categories pointing to a range of cellular, biological, and metabolic processes and emphasizing the intricate nature of psychiatric disorders. Such findings illustrate the formidable challenge of identifying and validating clinically useful metabolomic biomarkers and underscore the necessity for further research that encompasses various validation methodologies.

Keywords: Biomarkers; Metabolomics; Psychiatric disorders; Validation.

Fig. 1

The most prevalent psychiatric disorders worldwide according to World Health Organization (WHO) (World Health Organization 2022)

Fig. 2

Schematic representation of biomarker validation, arranged sequentially from the least to the most stringent type

Fig. 3

Distribution of validation metabolomic studies according to investigated psychiatric disorder (A) and according to the type of validation in each investigated psychiatric disorder (B)

Fig. 4

Total number of participants included in reviewed metabolomic validation studies divided per sample set (training or test sets) for each psychiatric disorder. The Case 1 group represents the group of patients with corresponding disorder, the Control group consists of healthy control subjects, while Case 2 group includes individuals with MDD who were studied alongside those with BD

Fig. 5

Percentage of the most common validated metabolic classes in the studied psychiatric disorders

Fig. 6

Simplified schematic representation of altered metabolic pathways in psychiatric disorders

Fig. 7

Summarized heat map of analyzed metabolites and their validation level. “I” = internal validation; “T” = temporal validation; “E” = external validation