INTRODUCTION
Major depressive disorder (MDD) is a severe condition that causes significant changes in emotional, cognitive, and neurovegetative functions, and its prevalence continues to increase annually [1]. Currently, the diagnosis of MDD is primarily based on clinical symptoms, following the Diagnostic and Statistical Manual of Mental Disorders (DSM) published by the American Psychological Association, or the International Classification of Diseases (ICD) system published by the WHO [2]. These diagnostic systems rely more on subjective reports and objective observations of individuals with MDD than on a diagnostic system based on the causes of the disease or on biological, chemical, and physical changes within the body, which are commonly used in most other medical fields outside psychiatry [3]. As a result, individuals with MDD, who may have various causes, end up forming a heterogeneous group, leading to diverse treatment outcomes for the same treatment [4,5]. To solve these problems, research in various fields has recently been conducted to identify the cause of MDD and find biomarkers for diagnosis [6-8].
The causes of MDD have not yet been clearly identified; however, current knowledge suggests that a combination of biological, psychological, and environmental factors contribute to the development of MDD [9]. Among the biological causes, monoamine, hypothalamic-pituitary-adrenal (HPA) axis, neuroendocrine, neuroanatomical, and genetic hypotheses have been proposed [10-12]. In the investigation of these causes, research is actively being conducted to identify biomarkers that could be integrated into the diagnostic process. These biomarkers can be used to objectively measure and evaluate normal biological processes (BPs) and pathological conditions [13]. Biomarker studies on MDD have reported to date include proteomic markers related to neurotransmitters, growth factors (e.g., BDNF), inflammatory markers (e.g., cytokines), metabolic and oxidative stress factors, endocrine markers, and biomarkers using functional magnetic resonance imaging (fMRI) and quantitative electroencephalography (qEEG) [14,15]. Although various potential biomarkers are being studied and proposed, it is known that there is still no biomarker that has been clearly identified with a consensus.
Among the various potential biomarker areas of MDD, the neuroimmunological theory states that since Smith [16] first introduced the concept that depression can be caused by the activation of immune cells in the macrophage theory of depression in 1991, evidence has been accumulating over the past 30 years that abnormalities in immune-inflammatory pathways and the activation of cell-mediated immunity are important pathophysiological pathways in the development of MDD [17]. The neuroimmunology hypothesis states that the pathogenesis of MDD and changes in cytokines are causal, suggesting that external and internal stress induce an imbalance of cytokine, which plays an important role in the expression and persistence of depressive symptoms in vulnerable individuals [18]. According to the cytokine hypothesis, the relationship between MDD and cytokines is explained based on research showing that cytokine injection induces depressive-like symptoms, an increase in cytokines is observed in patients with depression, and cytokines trigger the HPA axis and catecholamines that are closely related to MDD. Additionally, some studies have suggested that antidepressants suppress cytokine release from immune cells, further supporting the relationship between MDD and cytokines [12,19]. A meta-analysis of 82 related studies reported that levels of interleukin 6 (IL-6), tumor necrosis factor α (TNF-α), IL-10, soluble IL-2 receptor, C-C chemokine ligand 2, IL-13, IL-12, and IL-18 were higher in patients with MDD [20]. Another study compared cytokine levels in patients with MDD before and after antidepressant treatment, showing that responders had lower IL-8 levels than non-responders, and antidepressant treatment significantly reduced TNF-a levels only in responders [21]. A recent study indicated that peripheral inflammation may predict the onset of depression. This study showed that children with higher serum IL-6 levels at the age of 9 years had a 10% higher risk of developing MDD than the general population or children with lower IL-6 levels [22]. However, whether changes in specific cytokines can be used to definitively diagnose or predict depression remains unclear.
Early prediction of MDD using inflammatory markers, such as cytokines, can lead to more effective MDD treatment and prevention strategies, including proactive interventions and rapid response after onset. Previous studies have measured a small number of cytokines (less than 10) related to MDD, while this study analyzed a wider range of inflammatory cytokines simultaneously, including those that have not been extensively studied.
This study was conducted in two phases. First, potential biomarkers were identified at the transcriptome level using publicly available datasets. Second, these candidate biomarkers were validated at the protein level in clinical samples collected from patients with MDD. This multi-step approach enhances the reliability of our findings, supporting early detection, targeted prevention, and objective monitoring of disease progression alongside clinical evaluation.
METHODS
Overall study design
This study employed a two-step design. In the first phase, candidate biomarkers for MDD were identified through transcriptomic analysis using publicly available datasets. In the second phase, these selected candidate biomarkers were validated at the protein level using clinical samples collected from patients with MDD.
Data acquisition
The GSE39653 and GSE19738 datasets, both of which consist of microarray data obtained from peripheral blood mononuclear cells, were acquired from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The GSE39653 dataset includes 53 samples, comprising eight from bipolar disorder (BD), 24 healthy controls (HCs), and 21 patients with MDD. In this study, BD samples were excluded, leaving 45 samples (24 HC and 21 MDD) for analysis. The GSE19738 dataset contains 132 samples, equally divided into 66 HC and 66 MDD samples, all of which were used in this study.
Identification of differentially expressed genes
Differentially expressed gene (DEGs) between patients with MDD and HCs were identified using the GEO2R tool, which is an online interface based on the limma package for differential gene expression analysis. Given the relatively small sample size, we applied an adjusted p<0.05 and |log2foldchange| ≥0.1 as the threshold for identifying DEGs. To account for multiple comparisons, the Benjamini-Hochberg correction was applied to control the false discovery rate. These statistical criteria have been commonly used in similar studies analyzing small cohorts, including our previous research [23,24]. Common DEGs were identified by comparing the DEGs obtained from both datasets, and the overlapped genes were visualized using a Venn diagram. Venn diagrams were generated using the Venn package in R (version 1.12). DEG analysis results for each dataset were visualized using volcano plots, created using the ggplot2 package (version 3.5.1), available from the CRAN repository (https://cran.r-project.org/package=ggplot2).
Functional enrichment analysis of DEGs
To gain insight into the biological significance of the identified DEGs, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed. GO analysis classified genes into BP, cellular component (CC), and molecular function (MF) categories, while KEGG analysis identified key molecular pathways associated with MDD. This approach helps uncover the functional roles of dysregulated genes, offering a broader biological interpretation rather than focusing on individual genes. Both upregulated and downregulated DEGs were analyzed separately to distinguish activated and suppressed pathways. Enrichment analysis was conducted using the clusterProfiler package (version 4.12.2) in R Bioconductor, with p<0.05 as the significance threshold, and Benjamini-Hochberg correction was applied to control for multiple comparisons. While GO and KEGG analyses provide valuable insights into gene function and disease-related pathways, they are highly dependent on existing databases, which may lead to overrepresentation or underrepresentation of certain pathways due to database biases. Additionally, GO terms can exhibit high redundancy, making it challenging to extract biologically meaningful information. Despite these limitations, this approach remains one of the most widely used methods for functional annotation in transcriptomics and provides a foundation for further experimental validation of candidate genes and pathways.
Protein-protein interaction network construction and hub genes selection
To investigate the interactions among DEGs at the protein level, a protein-protein interaction (PPI) network was constructed using the STRING database (v11.5, https://string-db.org), which integrates experimentally validated and computationally predicted protein interactions. The network was visualized and analyzed using Cytoscape (version 3.9.1), and key gene clusters were identified using the MCODE plugin with predefined clustering criteria (network scoring degree >2, node score cut-off=0.2, k-core=2, and max depth=100) [25]. By identifying hub genes within the network, this approach helps pinpoint genes that may play a central role in MDD-related immune and inflammatory processes. While PPI network analysis is a powerful tool for understanding molecular interactions, it is not without limitations. The STRING database incorporates predicted interactions, which can lead to false positives, potentially skewing biological interpretations. Additionally, MCODE clustering prioritizes highly connected gene groups, meaning that biologically relevant but less interconnected genes may be overlooked. Despite these limitations, PPI network analysis remains a valuable approach for prioritizing key genes for further investigation and can complement functional enrichment analysis to reveal underlying molecular mechanisms in disease pathology.
Participants in blood analysis for cytokine microarray
Blood samples from 30 participants were analyzed. Of these, 20 were in the MDD patient group and 10 were in the HC group, which had not been diagnosed with psychiatric disorders. The patient group consisted of males and females aged 13-65 years who were diagnosed with MDD by a psychiatrist according to the diagnostic criteria of DSM-5. Blood samples were stored in the Keimyung University Dongsan Medical Center Biobank after being informed of the research purpose and providing informed consent for storage. Blood samples were collected during the active phase of depressive symptoms at the beginning of inpatient treatment. The 10 individuals in the control group gave their consent after receiving an explanation of the purpose of the study. Participants under 19 years of age also provided additional parental consent. Plasma was isolated from the blood samples collected from all participants using anticoagulant-containing tubes. The exclusion criteria included a history of infectious diseases, tumors, autoimmune diseases, brain disorders, or other medical conditions that could affect the study results, individuals with bleeding disorders unsuitable for blood collection, those taking anticoagulants (e.g., warfarin and aspirin), and individuals with intellectual disabilities or cognitive impairments. This study was approved by the Institutional Review Board (IRB) of Keimyung University Dongsan Medical Center (IRB No. 2022-11-077).
Proteomic analysis
For proteomic analysis, plasma was collected from 30 patients with depressive disorder at the time of diagnosis and stored at -80°C. The samples were sent to Olink Proteomics (Watertown) and analyzed using the Olink Target 96 Inflammation Panel. This panel was selected due to its unique inclusion of various inflammatory cytokines. The samples were processed using the Olink Target platform, which comprises a proximity extension assay (PEA) with next-generation sequencing (NGS). Potential protein biomarkers were evaluated using PEA as described on the manufacturer’s website (https://www.olink.com). The manufacturer’s website provides an extensive list of potential protein biomarkers that comprises 92 entries. Furthermore, the website offers comprehensive assay validation data, including details on the detection limit, lower and upper limits of quantification, and within- and between-run precision coefficients of variation. The final assay protein concentration output is presented in terms of normalized protein expression (NPX) values. NPX is a logarithmic unit on a base-two scale, in which elevated NPX values correspond to higher protein concentrations.
Predictive analysis using logistic regression
To evaluate the predictive power of potential biomarkers, the GSE260603 dataset was randomly divided into a training set and a validation set in an 8:2 ratio using the “caret” package in R [26]. Genes associated with the prediction of MDD onset were identified by univariate and multivariate logistic regression analyses facilitated by the R packages “glmnet” and “pROC.” [27,28] Logistic regression was selected because it estimates the probability that the data belong to a specific category, provides a value between 0 and 1, and classifies the data into the most likely category based on the estimated probability. This method is widely used for the predictive modeling of categorical data. Furthermore, to assess the discriminative ability of the gene combinations, receiver operating characteristic (ROC) curves were conducted and the area under the ROC curve was calculated for comparison.
Statistical analysis
To compare gene expression levels between patients with MDD and HCs, t-tests were performed using methods adjusted based on normality and homogeneity of variance. To compare the demographic and clinical variables between the MDD and HC groups, we used statistical methods that accounted for the characteristics of each variable. Due to the small sample size and potential for low expected frequencies, Fisher’s exact test was used to compare the sex proportions between the two groups. For age comparison, we first assessed the normality of age data using the Shapiro-Wilk test, and as both groups followed a normal distribution (p>0.05), a two-sample t-test was performed to compare the mean ages. The homogeneity of variances was also verified using the F-test (p>0.05). In the case of Hamilton Depression Rating Scale (HDRS) scores, normality testing indicated that the HDRS scores of the control group did not follow a normal distribution (p<0.05); therefore, the nonparametric Mann-Whitney U test was used to compare HDRS scores between the two groups. Statistical significance was set at p<0.05. All statistical analyses and visualizations were performed with the R software (version 4.1; R Foundation for Statistical Computing).
RESULTS
Identification of DEGs
Using datasets obtained from GEO, we identified DEGs associated with MDD. In the GSE39653 dataset, 1,815 DEGs were identified, of which 922 were upregulated and 893 were downregulated. In the GSE19738 dataset, 631 DEGs were identified, of which 471 were upregulated and 160 were downregulated. We identified 66 common DEGs by examining the overlap between the two datasets (Figure 1A). These common DEGs were considered potential biomarkers related to MDD and subjected to further analyses. The volcano plots that illustrate the DEGs for each dataset are shown in Figure 1B and C. To explore the biological significance of the identified DEGs, we performed GO and KEGG pathway enrichment analyses for each dataset. The results are presented in Supplementary Figures 1 and 2.
Functional enrichment analysis of common DEGs
Functional enrichment analysis was performed to determine the roles and signaling pathways associated with the common DEGs identified. GO analysis of common DEGs revealed significant terms, including leukocyte migration, cell chemotaxis, cytokine-mediated signaling pathway, external side of the plasma membrane, and receptor ligand activity signaling receptor activator activity (Figure 2A). KEGG analysis highlighted significant pathways, such as cytokine-cytokine receptor interaction, viral protein interaction with cytokines and cytokine receptors, and the chemokine signaling pathway (Figure 2B).
PPI network construction and hub genes selection
PPI analysis was performed to explore the interactions among the common DEGs at the protein level and to identify hub genes. Of these 66 common DEGs, 46 were identified and included in the PPI network (Figure 3A). GO analysis of the 46 genes within the PPI network identified significant terms, including cellular processes, regulation of BPs, and regulation of cellular processes (Figure 3B). KEGG pathway analysis of these genes indicated their significant involvement in cytokine-cytokine receptor interactions, viral protein interactions with cytokines and cytokine receptors, and chemokine signaling pathways (Figure 3C). These findings suggest that immune-related factors play a crucial role in the pathogenesis of MDD, particularly considering the prominent immune-related pathways identified in the GO and KEGG analyses.
Demographic and clinical data for the cytokine microarray
Of the 20 patients with MDD, six were males and 14 were females. The mean age±standard deviation was 27.55±14.05 years. The HC group had five males and five females, with a mean age of 24.30±11.70 years. In patients with MDD, the mean period from the onset of depressive symptoms to the diagnosis of MDD (duration of the episode) was 3.13±1.76 months. In the MDD patient group, the HDRS score measured at the start of inpatient treatment was 23.25±4.63. The HDRS score of the HC group, which was measured at the time of blood collection, was 1.80±2.30 (Table 1). Statistical analyses revealed that there was no significant difference in sex distribution between the MDD and control groups (Fisher’s exact test, p=0.425). Similarly, no significant difference was observed in age between the two groups (two-sample t-test, p=0.233). However, HDRS scores were significantly higher in the MDD group compared to the control group (Mann-Whitney U test, p<0.001), indicating a statistically significant difference in depression severity.
Experimental validation of common DEGs
From our bioinformatics analysis, we identified 46 potential biomarker genes related to MDD, most of which were associated with immune function. To validate these findings, we performed a cytokine array analysis using plasma samples from patients with MDD. The results indicated that 18 of the 46 genes were significantly upregulated in the plasma samples of patients with MDD (Figure 4), while the remaining 28 genes did not show statistically significant changes (Supplementary Figure 3). The genes showing significant differences were as follows: VEGFA, CXCL11, OSM, IL-2, CCL4, CD6, CCL11, TNFSF14, HGF, CD5, CCL3, SIRT2, CCL28, IL-33, IL-4, LIF, STAMBP, and ADA.
Evaluation of the predictive power of individual genes for MDD
The predictive powers of 18 potential biomarkers that were significantly upregulated at the protein level were assessed in an external MDD cohort. The analysis revealed that all biomarkers showed predictive ability, achieving an AUC≥0.6. The AUC values for individual genes were as follows: VEGFA (AUC=0.75), CXCL11 (AUC=0.81), OSM (AUC=0.96), IL-2 (AUC=0.84), CCL4 (AUC=0.74), CD6 (AUC=0.90), CCL11 (AUC=0.89), TNFSF14 (AUC=0.90), HGF (AUC=0.95), CD5 (AUC=0.88), CCL3 (AUC=0.72), SIRT2 (AUC=0.78), CCL28 (AUC=0.80), IL-33 (AUC=0.73), IL-4 (AUC=0.69), LIF (AUC=0.88), STAMBP (AUC=0.83), and ADA (AUC=0.87). Among these, OSM, CD6, TNFSF14, and HGF demonstrated strong predictive power (AUC>0.9) (Figure 5).
DISCUSSION
Recent studies have attempted to identify biomarkers for diagnosing and predicting the treatment response of various diseases using microarray and NGS technologies [29]. These whole-gene expression data have been studied extensively using differential expression analysis; however, there are concerns about the accuracy of the research results owing to factors such as clinical heterogeneity, data collection details, sample size, and biological variation. Consistent and repeated studies are necessary to improve the accuracy of these findings [30]. However, there are still relatively few studies that have used these methods in the context of depression. In this study, we analyzed various cytokines that have not been studied much before and attempted to discover high-confidence biomarkers by identifying and confirming potential biomarkers using proteomic analysis of plasma samples from patients with MDD. This will help us understand, diagnose, and treat MDD in the future.
In this study, we identified DEGs related to MDD using GEO datasets. As a result, a total of 66 DEGs were identified. Because these genes may interact with each other and have causal relationships, a functional enrichment analysis was performed to identify the roles and signaling pathways of these genes rather than simply confirming individual genes [31]. PPI analysis was performed to explore the interactions between common DEGs at the protein level and identify central genes. Finally, 46 genes were identified and the pathway analysis results of these genes mainly identified interactions with cytokines, cytokine receptors, and chemokine signaling pathways. The cytokine hypothesis, suggesting that cytokines contribute to the development of depression, is not independent of other established theories such as the monoamine hypothesis and hormone-related models [32,33]. According to the monoamine hypothesis, neurotransmitters like serotonin and norepinephrine (NE) play key roles in depression, and cytokines interact with these systems. Proinflammatory cytokines activate indoleamine 2,3-dioxygenase (IDO), which degrades tryptophan, the precursor of serotonin, thereby reducing serotonin availability [34]. Cytokines also affect neurotransmission by activating signaling pathways like p38 MAPK and ERK1/2, altering serotonin and NE reuptake [35]. Additionally, cytokines stimulate the secretion of corticotropin-releasing hormone (CRH) and adrenocorticotropic hormone (ACTH), leading to HPA axis hyperactivation and chronic cortisol elevation [36,37]. This disrupts neurotransmitter balance, amplifies stress responses, and increases depression risk. Chronic HPA axis activation also increases serotonin receptor numbers in key brain areas, such as the hippocampus, amygdala, and frontal cortex [36]. Therefore, the cytokines identified in this study can be proposed as immune-inflammatory factors in the etiology of MDD.
We also performed a proteomic analysis of inflammation-related cytokines in blood samples from patients with MDD and compared them with those of HCs. Among the 46 DEGs selected through the PPI network analysis, 18 showed significantly increased expression in MDD blood samples, while the remaining 28 did not show significant differences. Among the 18 cytokines identified in this study, those with the highest AUC values on the ROC curve were OSM (AUC=0.96), HGF (AUC=0.95), CD6 (AUC=0.90), TNFSF14 (AUC=0.90).
Oncostatin M (OSM) is a member of the IL-6 family of cytokines and has various functions, such as wound healing, liver regeneration, heart remodeling, pain, and inflammatory metabolism [38]. OSM is known to contribute significantly to autoimmune diseases, inflammatory diseases, and cancers that occur in various organs, and there is active research on its use in treatments [39]. In particular, OSM is maintained at a low level in the central nervous system (CNS) under normal conditions but increases in the pathological environment and can increase the expression of other cytokines, such as IL-6 [40]. Franzen et al. [41] reported that OSM levels differed in the cerebrospinal fluid samples from eight patients with MDD and seven HCs. Korhonen et al. [42] conducted a study of 110 participants, including patients with first episode psychosis and controls, and found increased levels of inflammatory cytokines, including OSM and IL-6. These findings are consistent with those of many previous studies, suggesting that inflammatory cytokines play an important role in the onset of MDD. Although specific research on OSM and MDD remains limited, these results suggest that OSM plays an important role in the development of MDD.
Hepatocyte growth factor (HGF) plays an essential role not only in the liver but also in various other organs by stimulating epithelial cell proliferation, motility, morphogenesis, and angiogenesis, thus providing critical signals for organ development. In the brain and CNS, HGF is crucial for cerebrovascular stability and neuronal regeneration and protection [43,44]. Furthermore, HGF is known to regulate immune responses, showing anti-inflammatory effects by inhibiting inflammatory cytokines such as IL-1, IL-6, and IFN-γ [45]. Meanwhile, Russo [46] observed lower levels of HGF in his study focusing on 26 depressed patients than in the control group, and this level was correlated with overall depressive behavior and depressive symptoms. Wakatsuki et al. [47] reported that mice with HGF suppression exhibited depressive and anxiety symptoms, which contrasts with the results of our study. However, Arnold et al. [48] found that plasma HGF levels increased as depressive symptoms increased in elderly aged 55-90 years and explained the opposite result of a previous study by suggesting that it may be a compensatory increase in response to elevated inflammatory cytokines. Many studies have reported not only an increase or decrease, but also an imbalance in inflammatory and anti-inflammatory cytokines [20,22,49]. However, a study of 30 depressed patients and 12 patients with MDD with borderline personality disorder reported no significant differences in HGF levels between the two groups [50,51]. Given these varied results under different conditions, future research is needed to confirm these findings with a larger sample size and standardized conditions.
CD6 is a signaling receptor expressed by lymphocytes and is involved in physiological processes such as lymphocyte activation and proliferation, and although the detailed biological mechanism is still unclear, it is known to play an important role in immune-mediated processes and disease pathogenesis [52]. CD6 is related to the development of autoimmune diseases, such as rheumatoid arthritis and multiple sclerosis, and CD6 antibodies are being studied as a treatment for these diseases [53,54]. In our study, CD6 levels were found to be elevated in patients with MDD compared to those in HCs; however, previous studies have rarely reported a relationship between CD6 and depression. Therefore, further studies on the role of CD6 in depression are warranted.
Tumor necrosis factor superfamily 14 (TNFSF14) is one of over 20 members of the TNF superfamily and plays an important role in cell apoptosis, proliferation, activation, and differentiation, and acts as a cytokine that regulates immunity and inflammatory responses [55]. TNFSF14 contributes to the pathogenesis of inflammatory diseases of the brain, mucosa, liver, joints, and vascular tissues, as well as autoimmune diseases and cancer [56,57]. In a study by Zhang et al. [58] involving 40 patients with MDD and 40 controls, TNFSF14 levels differed in patients with MDD, similar to the results of our study. In contrast, Yang et al. [59] conducted a study on 15 adolescent patients with depression and 15 controls and reported that TNFSF14 was downregulated in adolescent patients with depression, which is contrary to our findings. Many studies on cytokines in MDD, including those mentioned above, have reported an imbalance in cytokines.
Multiple studies have observed changes in cytokine levels among patients with depression, but it remains unclear whether these changes are causative or consequential [17,18]. Administration of cytokine can induce depressive-like symptoms, suggesting a causal role; however, chronic stress can activate the HPA axis, autonomic nervous system, and immune system, indicating that these changes may be a consequence of depression [12,60]. Depression is a heterogeneous disorder with similar symptoms influenced by various factors, including neurotransmitter imbalances, hormonal dysregulation, environmental stress and psychosocial elements [61]. Longitudinal studies in healthy individuals could help clarify this relationship in the future. Further research is needed to elucidate the exact mechanisms underlying these dysregulations.
Previous meta-analyses studying inflammatory cytokines in depression reported that IL-2, IL-6, and IL-10 were elevated in depressed patients, while IL-4 and IFN-γ were reduced [20]. In this study, IL-2 was also significantly elevated in patients with depression, consistent with these findings, but there were no significant differences for IL-6, IL-10, and IFN-γ. One study suggested that there were no overall differences in IFN-γ between patients and controls, but sensitivity analysis showed various results potentially influenced by clinical factors such as smoking and high body mass index (BMI) [62]. In this study, IL-4 levels were significantly elevated in depressed patients, contrary to recent meta-analysis studies [62]. However, some studies comparing cytokines between patients with depression and HCs reported no significant differences in serum IL-4 levels or found increased levels in patient groups, as observed in our study, highlighting the need for further research [63-66].
Limitations
There are some limitations to this study. First, this study analyzed data only from the publicly available GEO dataset; therefore, it cannot be said to represent all previous research findings and there may be some differences from previous results. Second, the proteomic analysis of blood samples from patients with MDD had a small sample size and was carried out in a single center due to the high costs of the analysis and the limitations of the research process. Based on these results, future studies should investigate larger sample sizes to assess the potential of a more focused selection of cytokines as biomarkers. Third, the variables that could affect cytokine levels were not sufficiently controlled. Due to the retrospective nature of this study, we were unable to fully control for the types of medications administered to all participants, which may have influenced the study outcomes. Additionally, although all participants were primarily diagnosed with MDD, some patients also presented with comorbid psychiatric conditions, such as anxiety disorders. These factors should be taken into consideration when interpreting the results. Also, some antidepressants are known to affect cytokine levels [67,68]; however, since this was a retrospective study, patients were treated with different medications according to a psychiatrist’s judgment. Other variables can affect cytokine levels, such as smoking, obesity, and excessive alcohol consumption [62,69], and future studies are needed to identify and systematically analyze various related clinical variables.
Conclusion
Taken together, by analyzing results from public datasets related to previous MDD studies and blood samples from patients with MDD and HCs, we observed elevated levels of inflammatory cytokines, such as plasma OSM, HGF, CD6, and TNFSF14, in patients with MDD, suggesting the importance of altered cytokines as potential biomarkers that could be related to important pathophysiological aspects of MDD. These findings may contribute to the development of effective strategies for the prevention and treatment of MDD. Future studies should involve systematic and prospective research to control for potential confounding factors and further clarify the relationship between inflammatory markers and depression.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement1
Print
