Bipolar disorder (BD) and major depressive disorder (MDD) are two of the most prevalent mood disorders, usually associated with significant disability and mortality. Etiologically related, but clinically distinct, they share important clinical overlapping characteristics, which result in higher rates of misdiagnosis. The frequent occurrence of depressive episodes and the later onset of mania in BD subjects often delay a proper diagnosis for years, resulting in greater severity of symptoms, impaired psychosocial functioning, treatment resistance, and higher suicidality. Such delay also leads to a higher number of relapses and hospitalizations, with increased direct and indirect costs associated with the treatment and management of these disorders [Reference McIntyre, Berk, Brietzke, Goldstein, López-Jaramillo and Kessing1].

The process of diagnosis is quite complex in psychiatry and effective recommendations to improve the differential diagnosis between BD and MDD are lacking. This is of particular relevance, considering that available classification systems are not able to catch the complexity of these mental disorders. In fact, the definition of clear boundaries between psychopathology and the individual variability in behavioral manifestations, as well as between clinical psychiatric entities, is tricky and may result in frequent co-occurrence and heterogeneity within disorders and diagnostic instability over time. This has led to the idea that a categorial approach to mental disorders while representing a milestone in the harmonization of psychiatric diagnoses across different countries, cultures and clinicians, may be too simplistic [Reference Misiak, Samochowiec, Kowalski, Gaebel, Bassetti and Chan2].

Several efforts have been made to overcome the categorical approach to mental illness, one being represented by the Hierarchical Taxonomy of Psychopathology (HiTOP) project, that points to the development of a consensus dimensional classification more clinically informative than the traditional diagnostic systems [Reference DeYoung, Blain, Latzman, Grazioplene, Haltigan and Kotov3]. HiTOP aims at reducing heterogeneity by defining empirically coherent dimensions and accommodating the widespread comorbidity and poor treatment specificity associated with categorical diagnoses. This is achieved by organizing clinical phenotypes into transdiagnostic spectra and superspectra, which include both psychopathological and neurobiological elements [Reference DeYoung, Blain, Latzman, Grazioplene, Haltigan and Kotov3].

Another element hindering the achievement of a proper diagnosis relies on the “human factor.” For example, partial or inaccurate psychiatric history reporting and/or the inability to have complete access to previous medical records, as well as the possible inexperience of clinicians, may all represent a source of bias. It is clear that the classical clinical approach alone may be unable to pinpoint objective criteria to formulate a proper and reliable diagnosis, calling for unbiased biological indicators. Such biomarkers and their crosstalk may, therefore, represent the biosignature of the disease, reflecting the fundamental pathophysiologic mechanisms that distinguish MDD from BD [Reference Abi-Dargham, Moeller, Ali, DeLorenzo, Domschke, Horga, Jutla, Kotov, Paulus, Rubio, Sanacora, Veenstra-VanderWeele and Krystal4]. Once identified, such biosignature may guide the diagnosis also offering biological targets for individualized treatment and the identification of novel interventions for BD and MDD.

Affective disorders are characterized by multifactorial etiopathogenesis and complex pathophysiology that result from the interplay between genetic background and epigenetic phenomena occurring during early developmental phases [Reference Sampogna, Fiorillo, Luciano, Del Vecchio, Steardo and Pocai5]. Moreover, throughout their entire life, individuals will continue to challenge and shape the neural, immune, endocrine, and enteric systems, by actively interacting with the environment ultimately contributing to the heterogeneity characterizing the presentation of clinical symptoms of BD and MDD. Indeed, growing evidence suggests substantial dysregulations of immunological/inflammatory processes and increased oxidative stress (OS) in individuals with affective disorders; these same dysregulations have also been listed in the pathophysiology of metabolic disorders [Reference Sampogna, Fiorillo, Luciano, Del Vecchio, Steardo and Pocai5].

Not surprisingly, affective disorders are among the psychiatric illnesses most frequently associated with a reduced life expectancy, compared to the general population, mainly as a result of cardiovascular or metabolic comorbidities [Reference Sampogna, Fiorillo, Luciano, Del Vecchio, Steardo and Pocai5]. Indeed, those experiencing affective disorders with physical comorbidities usually present greater severity of psychiatric symptoms, higher rates of resistance to pharmacological treatments, lower recovery rates, worse long-term outcomes, and reduced psychosocial functioning and quality of life. On the other hand, people suffering from cardiometabolic disorders appear at an increased risk of developing affective disorders throughout life [Reference Sampogna, Fiorillo, Luciano, Del Vecchio, Steardo and Pocai5]. Taken together, current studies seem to converge on the hypothesis that metabolic and affective disorders may reinforce one another in a vicious cycle relying upon a bidirectional brain – body communication based – at least in part – on immune-inflammatory and OS mediators. Such hypothesis is also strengthened by preclinical research providing evidence for shared common pathways in utero, with prenatal stressors accounting for programmed fetal liability [Reference Musillo, Berry and Cirulli6]. In fact, it has been recently suggested that, regardless of the specific nature of maternal stress (either metabolic or psychosocial), OS might be one of the main triggers disrupting fetal development [Reference Musillo, Berry and Cirulli6]. Indeed, the mammalian brain is characterized by poor antioxidant defenses, high metabolic rate, and reduced capacity for cellular regeneration being susceptible to OS and inflammatory insults [Reference Berry and Cirulli7]. Evidence supporting the main role of OS in psychiatric disorders comes from studies on the positive role of antioxidant compounds in a wide range of clinical conditions including MDD and BD. Worth mentioning, specific reactive oxygen species (H₂O₂), produced by mitochondria, reinforce insulin signaling promoting metabolic disorders. By contrast, animal models with genetically reduced OS are characterized by altered emotional behavior [Reference Berry and Cirulli7].

From the above-mentioned evidence, it appears clear that changes in the metabolic and redox set-point might play a key role in the pathophysiology and/or in the etiopathogenesis of mood disorders and that immunometabolic pathways may represent a main target in the search for biomarkers. The rationale for searching peripheral signatures of brain disorders relies on the notion, rapidly evolving, that affective disorders are characterized by disrupted brain – body interactions. This ultimately means that a single biomarker, as is the case of some kind of cancers, can hardly represent the specific biosignature of a psychiatric disorder.

Recent efforts to identify disease-specific biomarkers have provided evidence for distinct immune patterns and non-enzymatic antioxidant alterations in BD and MDD. In particular, Gong and co-workers [Reference Gong, Lu, Kang, Feng, Zhang and Sun8] found that levels of circulating non-enzymatic antioxidants might discriminate individuals with BD from those with MDD, while Poletti and colleagues [Reference P Poletti, Vai, Mazza, Zanardi, Lorenzi and Calesella9], have shown that BD and MDD subjects are characterized by disease-specific immune profiles. Preliminary evidence shows that BD is characterized by an altered metabolic and inflammatory profile during the active phases of the disease [Reference Di Vincenzo, Luciano, Collacchi, De Felice, Giona and Piegari10]. This study is specifically aimed at integrating the immunometabolic and neuroendocrine peripheral biomarkers through a machine-learning approach in the search of a reliable and highly predictive biosignature of the disease. The alterations above described are positively correlated with anxiety, mania, depressive symptoms and the cumulative score for the severity of physical comorbidities. In the same study, when compared to healthy controls, individuals with BD present also a specific immune profile confirming that a selective BD biosignature characterizes the euthymic phase. We are currently extending these findings to include MDD subjects, while refining the biomarkers tested.

At the current level of knowledge, imaging studies represent a valuable non-invasive strategy to complement the study of peripheral biomarkers enhancing the ability to perform a differential diagnosis. However, although promising, this approach is costly and time-consuming, which hampers its massive use in preclinical research. Given the high degree of central and peripheral communication in mood disorders, a promising approach might be to study extracellular vesicles (EV): released from almost all types of cells, including neurons, they may be an additional innovative mean of having direct access to brain-derived molecular markers, and possibly representing a diagnostic implementation of a peripheral, disease-specific biosignature. EV (including exosomes) can cross the blood – brain barrier from both directions and carry proteins, metabolites, and nucleic acids, and once their cargo is delivered into recipient cells, they can alter their immunometabolic response. Although it is still tricky to identify the brain origin of exosomes, this approach is promising in that it provides a “window-into-the-brain” and holds great potential for future advancements in the comprehension of molecular mechanisms underlying psychiatric disorders, their course and outcomes.

Biomarkers can be considered clinically significant when they hold a relevant diagnostic, prognostic, or predictive value. Major advancements in identifying a disease-specific biosignature might derive not only from the characterization of selected biological markers but also from a deep understanding of how they interact with psychological and contextual factors that might affect their expression, ultimately leading to meaningful changes for individuals and the whole health system.