Oxidative stress biomarkers across mood episodes in bipolar disorder: potential roles of TrxR1 and PRDX1 Authors: Onur Gökçen, Kader Semra Karataş, Merve Akkuş, Feyza Dönmez, Ayşe Koçak Sezgin, Meliha Koldemir Gündüz, Güllü Kaymak Published: 2026 ABSTRACT AbstractAimOxidative stress is considered to play a role in the etiopathogenesis of bipolar disorder. This study aimed to compare oxidative stress indicators among bipolar disorder patients in different mood states and healthy controls.MethodThe sample included 101 bipolar patients (33 euthymic, 36 depressive, and 32 manic) and 29 healthy controls. Total antioxidant status (TAS), total oxidation status (TOS), superoxide dismutase (SOD), glutathione (GSH), thioredoxin reductase 1 (TrxR1), and peroxiredoxin 1 (PRDX1) levels were measured, along with SOD2 rs4880 and GPX3 rs3792797 single nucleotide polymorphisms (SNPs).ResultsCompared to controls, bipolar patients had higher TAS (p = 0.031), TrxR1 (p < 0.001), and PRDX1 (p < 0.001) levels, and lower GSH levels (p < 0.001). No significant group differences were found for TOS (p = 0.776), SOD (p = 0.086), or OSI (p = 0.312). Among mood states, TAS, GSH, TrxR1, and PRDX1 levels did not differ significantly. TrxR1 and PRDX1 showed significant diagnostic performance in distinguishing bipolar disorder (p < 0.001 for both). While GPX3 rs3792797 SNP did not differ between patients and controls, a significant difference was found for SOD2 rs4880, with AA allele carriers exhibiting higher SOD levels.ConclusionThis study is among the few investigating oxidative stress markers across different mood episodes in bipolar disorder and the first to examine TrxR1 and PRDX1. Findings suggest TrxR1 and PRDX1 as potential novel biomarkers in the pathophysiology of bipolar disorder.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12888-026-08179-7.Keywords: Bipolar disorder, Oxidative stress, Superoxide dismutase, Glutathione, Thioredoxin reductase 1, Peroxiredoxin 1 Introduction Bipolar disorder (BD) is a chronic psychiatric disorder characterized by recurrent mood episodes, which create a significant socioeconomic burden on society [29]. BD has been associated with a high rate of disability and suicide [46]. Although the exact cause of BD is not completely known, it is considered to occur due to the combination of genetic, immunological, and environmental factors [45]. Among these mechanisms, oxidative stress has been suggested to play a particularly important role in the pathophysiology of BD [62]. Oxygen-derived free radicals with unpaired electrons are called reactive oxygen species (ROS) [52]. Oxidative stress results from an imbalance between ROS production and their elimination by the antioxidant system. ROS induce lipid peroxidation in membranes, especially in brain tissue, which is more sensitive to oxidative stress, and can damage neurons [12]. Oxidative stress is known to contribute to reduced neurogenesis and increased apoptosis and neurodegeneration in BD [32]. Markers produced from ROS effects and antioxidant enzyme activities have been used in various studies to determine oxidative stress status. Total antioxidant status (TAS), total oxidant status (TOS), and the oxidative stress index (OSI) are the main indicators reflecting the balance between oxidation and antioxidation in the body. TAS represents the total activity of all antioxidants, while TOS is used as an indicator of ROS. OSI is calculated as the ratio of TOS to TAS and indicates the level of oxidative stress [20]. There are conflicting data in studies conducted on BD [3, 21, 42, 54]. Superoxide dismutase (SOD) and glutathione (GSH) are the main antioxidant defense systems involved in combating oxidative damage. SOD oxidizes a superoxide radical to oxygen molecule and reduces another superoxide radical to hydrogen peroxide (H2O2), a less reactive molecule [24]. GSH, the main antioxidant of the brain, is converted to oxidized glutathione (GSSG) by the enzyme glutathione peroxidase to reduce hydrogen peroxide levels [28]. Another important antioxidant system in the central nervous system is the thioredoxin system. Thioredoxin reductase reduces thioredoxin by transferring electrons, and the reduced thioredoxin subsequently reactivates oxidized peroxiredoxin. In this way, peroxiredoxin continues to reduce reactive oxygen species such as hydrogen peroxide. This cycle is a critical defense mechanism that protects the cell from oxidative damage [9]. There are conflicting results in studies comparing SOD levels in patients with bipolar disorder (BD) and healthy controls [38]. Although many studies have demonstrated reduced GSH levels in BD, some have reported no significant differences relative to healthy controls [32, 51]. A few studies comparing GSH and SOD levels in patients during different mood episodes have also shown varying results [32, 40, 56]. On the other hand, the single nucleotide polymorphisms (SNPs) superoxide dismutase 2 (SOD2) rs4880 and glutathione peroxidase 3 (GPX3) rs3792797 have been associated with an increased risk for BD [15, 25]. Previous studies have suggested that genetically determined impairments in oxidative defense mechanisms may contribute to the pathophysiology of BD. In this context, SOD2 rs4880 and GPX3 rs3792797 have been linked to neuroanatomical alterations and cognitive effects in individuals with BD, supporting the notion that variations in antioxidant-related genes may influence both biological vulnerability and clinical manifestations of the disorder [68]. Thioredoxin reductase 1 (TrxR1) and peroxiredoxin 1 (PRDX1) levels have been linked to many pathological conditions, and some studies suggest that oxidative stress may be a potential marker for psychiatric disorders [5, 17, 35, 37, 39, 47, 67]. However, to our knowledge, there are no studies yet on PRDX1 and TrxR1 in bipolar disorder. Previous studies investigating oxidative stress in bipolar disorder have yielded inconsistent findings, which may be related to differences in clinical characteristics, mood states, medication exposure, and the specific biomarkers examined. Therefore, in the present study, oxidative stress was approached as a multilayered biological process rather than through a single marker. TAS, TOS, and OSI were selected to reflect the global oxidant-antioxidant balance; SOD and GSH were included as representative components of classical enzymatic and non-enzymatic antioxidant defense systems, given their central role in cellular redox homeostasis; and TrxR1 and PRDX1 were examined to evaluate the thioredoxin/peroxiredoxin system, a more specific thiol-based redox pathway that has been relatively underexplored in bipolar disorder. By assessing these parameters together within the same sample, we aimed to obtain a more integrated picture of oxidative stress in bipolar disorder. Accordingly, the aim of this study was to compare oxidative stress indicators among patients with bipolar disorder in depressive, manic, and euthymic states and healthy controls, and to explore whether different levels of redox regulation vary across mood episodes. In addition, SOD2 rs4880 and GPX3 rs3792797 polymorphisms were included as exploratory genetic variables to examine whether selected antioxidant-related variants might be associated with biomarker levels. Methods Subjects All participants were recruited from the Kütahya Health Sciences University Medical Faculty Department of Psychiatry outpatients clinic. Ethical approval was obtained from the Independent Ethics Committee of Kütahya Health Sciences University (Decision No: 2023/06–19, dated May 11, 2023). This study was carried out in compliance with the ethical guidelines set by the responsible committee on human experimentation and the Declaration of Helsinki (2013). Written informed consent was obtained from all participants. A total of 101 patients diagnosed with bipolar disorder (33 in euthymic episode, 36 in depression and 32 in mania) and 29 healthy controls were included in the study. The diagnosis of bipolar disorder was confirmed according to the Structured Clinical Interview for DSM-V clinical version (SCID-CV) [18]. A structured clinical interview, clinical version for DSM-V (SCID-CV) was also performed in the control group to exclude psychiatric disorders. Participants were aged between 18 and 65 years. Exclusion Criteria for All Participants: Under 18 years of age or over 65 years of age. Under 18 years of age or over 65 years of age. Body mass index (BMI) over 30 kg/m². Body mass index (BMI) over 30 kg/m². History and treatment of systemic diseases such as hyperlipidemia, diabetes mellitus, coronary artery disease, hypertension, hypothyroidism, hyperthyroidism, acute infection, malignancy, neuroinflammatory or neurodegenerative diseases. History and treatment of systemic diseases such as hyperlipidemia, diabetes mellitus, coronary artery disease, hypertension, hypothyroidism, hyperthyroidism, acute infection, malignancy, neuroinflammatory or neurodegenerative diseases. Alcohol or substance abuse. Alcohol or substance abuse. Use of corticosteroids or other immunosuppressive drugs. Use of corticosteroids or other immunosuppressive drugs. Use of antioxidant supplements (e.g., vitamins, N-acetylcysteine, omega-3, or herbal compounds) or probiotics within the past six months. Use of antioxidant supplements (e.g., vitamins, N-acetylcysteine, omega-3, or herbal compounds) or probiotics within the past six months. Intellectual disabilities or severe neurological disorders. Intellectual disabilities or severe neurological disorders. Sociodemographic and clinical information of the participants was recorded using a form prepared by the researchers. This form was developed specifically for the present study and has not been previously published elsewhere. An English version of the form has been uploaded as a supplementary file and is cited accordingly in the manuscript. All treatments currently used by the participants were also recorded. The Hamilton Depression Rating Scale-17 (HDRS-17) was used to assess the severity of depressive symptoms, and the Young Mania Rating Scale (YMRS) was used to assess the severity of manic symptoms [4, 27, 34, 64]. Participants with HDRS-17 scores below 8 and YMRS scores below 6 were considered euthymic. Sample collection and processing of blood samples for oxidative stress parameters evaluation Blood samples of 5 ml were collected from each patient and control via venipuncture, utilizing tubes without anticoagulants or heparin. Serum was obtained through centrifugation at 3000 g for 5 min. TOS and TAS determination Kits (Rel Assay Diagnostics reagent; Mega Tıp, Gaziantep, Turkey) developed by Erel were used to determine TAS and TOS. The TAS level was measured using the Erel method. This method measures the antioxidative effect of the sample against potent free radical reactions initiated by produced hydroxyl radicals [19]. The results are presented as µmol Trolox Eq/L. The TOS value was measured using a novel method developed by Erel [20]. The intensity of color, measurable via spectrophotometry, correlates with the total quantity of oxidant molecules in the experimental sample. The assay was calibrated using hydrogen peroxide (H2O2), and results were reported as micromolar hydrogen peroxide equivalents per liter (µmol H2O2 Eq/L) via spectrophotometry [36]. All samples were analyzed in duplicate to ensure analytical precision. Oxidative stress index (OSI) was calculated as total oxidant status (TOS) divided by total antioxidant status (TAS). GSH level assay The GSH level was ascertained by employing the methodology outlined by Beutler et al. [7]. Initially, 200 µL of serum sample was diluted in 800 µl of phosphate buffer, and the initial absorbance (OD1) was recorded at 412 nm. Subsequently, 100 µl of Ellman’s reagent was introduced into the same tube, and the second absorbance value (OD2) was measured. SOD determination ELISA (Enzyme-Linked Immunosorbent Assay) kit ELK Biotechnology (Cat: ELK8777, Denver, CO 80202 USA) was used for serum SOD determination. All experiments were conducted according to the kit instructions. All samples were analyzed in duplicate to ensure analytical precision. Results were given U/L. Thioredoxin reductase 1 levels and peroxiredoxin-1 levels determination ELISA (Enzyme-Linked Immunosorbent Assay) kit ELK Biotechnology (Cat: ELK2000, Denver, CO 80202 USA) was used for serum thioredoxin reductase 1 levels determination. For the quantitative measurement of peroxiredoxin-1 levels determination in serum we used ELISA kit ELK Biotechnology (Cat: ELK3490, Denver, CO 80202 USA). All experiments were conducted according to the kit instructions. All samples were analyzed in duplicate to ensure analytical precision. Results were given ng/mL. Detection of single nucleotide polymorphisms in the SOD2 gene rs4880 and the GPX3 gene rs3792797 The SOD2 gene rs4880 and GPX3 Gene rs3792797 polymorphisms were determined using a TaqMan SNP Genotyping Assays (Applied Biosystems, Foster City, CA) in the qPCR. TaqMan polymerase chain reactions (PCR) 1 x ABI TaqMan Genotyping Master Mix, TaqMan 20 x probe mix and 20 ng DNA for a total volume of 10 uL were used to create a mixture for genotyping. The ABI StepOnePlus real-time PCR system (Applied Biosystems, Germany) was used to analyze the mixtures that were prepared for the assessment of the rs4880 and rs3792797 genotypes. Each analysis was conducted thrice. Statistical analysis All statistical analyses were performed using IBM SPSS version 25.0 (SPSS Inc., Chicago, Illinois, USA). Continuous variables are presented as mean ± SD in the tables. Categorical variables are presented as n (number) and % (percentage). When the continuous data of the study were analyzed in terms of normality assumptions, it was determined that the Kolmogorov-Smirnov values did not demonstrate a normal distribution because the values obtained from the Kolmogorov-Smirnov test were p < 0.05 and the Skewness and Kurtosis values, divided by their standard errors, were not within the threshold value range of ± 3.29 [44, 59]. The Mann-Whitney U test, one of the nonparametric tests for variables with two groups, and the Kruskal-Wallis H test for variables with three or more groups were applied. In case of a significant difference between the groups, Dunn’s test, a post-hoc test, was preferred to determine between which groups the significance lay. To address the risk of type I error due to multiple testing, a Bonferroni correction was applied across the 8 principal parameters examined in the study (TAS, TOS, SOD, GSH, TrxR1, PRDX1, SOD2 rs4880, and GPX3 rs3792797), and the corrected significance threshold was set at p < 0.00625. Because some genotype categories contained very small cell counts, comparisons of genotype distributions between bipolar disorder patients and healthy controls were performed using Fisher’s exact test with Monte Carlo simulated p-values rather than relying solely on asymptotic chi-square estimates. ROC analysis was performed to determine the predictive cut-off points of TrxR1 and PRDX1 values in differentiating bipolar disorder. p < 0.05 was considered statistically significant. No separate a priori sample size or power calculation was performed specifically for the genetic analyses. The final sample was determined by the number of eligible participants who were consecutively recruited during the study period; therefore, the SNP analyses were considered exploratory. Results The findings related to demographic and clinical characteristics of the participants No statistically significant difference was found between the bipolar and control groups in terms of age, smoking, BMI and gender (p:0.398 for smoking, p:0.579 for age, 0.297 for BMI, p:0.408 for gender). Of the participants, 25.4% were euthymic, 27.7% depression, 24.6% mania and 22.3% control group. 48.5% of the participants were female and 50% were smokers. It was determined that 33.7% of bipolar disorder patients used lithium, 52.5% used valproic acid, 3% used carbamazepine, 78.2% used antipsychotics and 22.8% used antidepressants (Table 1). Table 1The sociodemographic and clinical features of the patients with bipolar disorder and healthy controls (n = 130)VariablesStatistic Patients with BD 101 (%77.7) Euthymic33 (%25.4) Depression36 (%27.7) Mania32 (%24.6) HCs 29 (%22.3) Female 63 (%48.5) Smokers 65 (%50.0) Lithium 34 (% 33.7) Valproic acid 53 (% 52.5) Carbamazepine 3 (%3) Lamotrigine 12 (%11.9) Antipsychotics 79 (%78.2) Antidepressants 23 (%22.8) Age (years) 40.72 ± 12.35 Illness Duration (years) 13.6 ± 11.6 BMI (Kg/m 2 ) 26.84 ± 2.72Data expressed in mean ± SD or frequency (%)BD= Bipolar Disorder, HCs= Healthy Controls, BMI= Body mass index The sociodemographic and clinical features of the patients with bipolar disorder and healthy controls (n = 130) Data expressed in mean ± SD or frequency (%) BD= Bipolar Disorder, HCs= Healthy Controls, BMI= Body mass index The differences between groups in terms of oxidative stress variables A statistically significant difference was observed between the groups in terms of TAS (p = 0.031, Table 2). TAS levels were found to be higher in individuals with mania. Although DUNN’s post hoc test revealed no statistically significant differences between any two groups, a comparison of all patients diagnosed with bipolar disorder and the control group showed that TAS was significantly higher in individuals with bipolar disorder, according to the Mann-Whitney U test results (p = 0.003). Table 2A comparison of oxidative stress variables and scale scores between groupsEuthymic (n = 33)Depression (n = 36)Mania (n = 32)Healthy Controls(n = 29) p Post- HocHDRS-171.59 ± 1.2418.06 ± 5.787.25 ± 5.35-YMRS0.81 ± 1.124.17 ± 2.8519.88 ± 10.09-TAS1.38 ± 0.281.43 ± 0.291.47 ± 0.391.17 ± 0.32 0.031 -TOS0.94 ± 0.541.02 ± 0.511.01 ± 0.421.02 ± 0.480.776-OSI0.69 ± 0.390.76 ± 0.420.7 ± 0.230.98 ± 0.70.312-SOD474.54 ± 296.72405.39 ± 175.31343.72 ± 112.11362.53 ± 37.090.086-GSH0.10 ± 0.160.05 ± 0.040.05 ± 0.060.15 ± 0.04 < 0.001 1,2,3 < 4TrxR13.75 ± 1.63.33 ± 1.813.90 ± 2.130.83 ± 0.99 < 0.001 1,2,3 > 4PRDX15.49 ± 3.95.02 ± 3.866.46 ± 4.41.19 ± 0.47 < 0.001 1,2,3 > 4Data expressed in mean ± SD, CV = Kruskal-Wallis H test, Post-Hoc = Dunn’s test, p < 0.05, Group codes used in post hoc comparisons: 1 = euthymic bipolar disorder, 2 = bipolar depression, 3 = bipolar mania, and 4 = healthy controlsHDRS-17 = The Hamilton Depression Rating Scale-17; YMRS = The Young Mania Rating Scale A comparison of oxidative stress variables and scale scores between groups Data expressed in mean ± SD, CV = Kruskal-Wallis H test, Post-Hoc = Dunn’s test, p < 0.05, Group codes used in post hoc comparisons: 1 = euthymic bipolar disorder, 2 = bipolar depression, 3 = bipolar mania, and 4 = healthy controls HDRS-17 = The Hamilton Depression Rating Scale-17; YMRS = The Young Mania Rating Scale A statistically significant difference was observed between the groups in terms of GSH (p < 0.001, Table 2). GSH values of the control group were found to be higher. According to the results of DUNN’s post-hoc analysis, statistically significant differences were found between the euthymic group and the control group (p < 0.001), the depression group and the control group (p < 0.001), and the mania group and the control group (p < 0.001). A statistically significant difference was observed between the groups in terms of TrxR1 (p < 0.001, Table 2). TrxR1 values of the mania group were found to be higher. According to the results of DUNN’s post-hoc analysis, statistically significant differences were found between the euthymic group and the control group (p < 0.001), the depression group and the control group (p < 0.001), and the mania group and the control group (p < 0.001). A statistically significant difference was observed between the groups in terms of PRDX1 (p < 0.001, Table 2). PRDX1 values of the mania group were found to be higher. According to the results of DUNN’s post-hoc analysis, statistically significant differences were found between the euthymic group and the control group (p < 0.001), the depression group and the control group (p < 0.001), and the mania group and the control group (p < 0.001). To provide a more integrated view of the oxidative stress profile, additional exploratory Spearman correlation analyses were performed among the principal biochemical markers. TrxR1 showed a moderate positive correlation with PRDX1 (rho = 0.527, p < 0.001). PRDX1 was positively correlated with SOD (rho = 0.322, p < 0.001), whereas GSH was negatively correlated with both TrxR1 (rho = -0.339, p < 0.001) and PRDX1 (rho = -0.366, p < 0.001). TAS showed only weak correlations with TrxR1 and PRDX1, and no significant correlation with SOD or GSH. The effect of treatments on oxidative stress variables in bipolar patients When the results of bipolar patients were evaluated according to drug use, SOD values of lamotrigine users (p = 0.031) and SOD values of antidepressant users (p = 0.010) were found to be statistically significant and higher than those of non-users. No difference was found in terms of other treatments and values. SOD2 rs4880 and GPX3 rs3792797 SNPs Table 3 shows the results of the analyses comparing patients with a bipolar diagnosis and healthy controls in terms of SOD2 rs4880 and GPX3 rs3792797 SNPs. According to the results of this analysis, a statistically significant difference was observed between patients diagnosed with bipolar disorder and healthy controls regarding the SOD2 rs4880 genotypes (p < 0.001). Table 3Comparison of bipolar and control groups in terms of SOD2 rs4880 and GPX3 rs3792797 SNPsHCs(n = 29)Patients with BD(n = 96) p GPX3 rs3792797AA1 (11.1)8 (88.9)AC10 (28.6)25 (71.4)0.478CC17 (21.3)63 (78.8)AA9 (21.4)33 (78.6)SOD2 rs4880AG2 (4.3)45 (95.7) < 0.001 GG18 (50)18 (50)Fisher’s exact test, BD= Bipolar Disorder, HCs= Healthy Controls Comparison of bipolar and control groups in terms of SOD2 rs4880 and GPX3 rs3792797 SNPs Fisher’s exact test, BD= Bipolar Disorder, HCs= Healthy Controls To address the risk of type I error due to multiple testing, a Bonferroni correction was applied across the 8 principal parameters examined in the study (corrected significance threshold: p < 0.00625). After this correction, the differences in GSH, TrxR1, and PRDX1 levels, as well as the difference in SOD2 rs4880 genotype distribution, remained statistically significant, whereas the difference in TAS did not remain significant after correction. As shown in Table 4, all participants, including both the patient and control groups, were grouped according to their genotypes, and no statistically significant difference was observed between the GPX3 rs3792797 genotypes in terms of any parameter. However, a statistically significant difference was identified between the SOD2 rs4880 genotypes regarding SOD levels (p = 0.030). Participants with the AA allele exhibited higher SOD levels compared to those with AG and GG alleles. Table 4Comparison of SOD2 rs4880 and GPX3 rs3792797 genotypes in terms of oxidative stress variables n TASTOSOSISODGSHTrxR1PRDX1 GPX3 rs3792797 CC801.33 ± 0.351.03 ± 0.520.83 ± 0.54412.4 ± 222.660.09 ± 0.113.14 ± 2.264.85 ± 4.19AA + AC441.42 ± 0.320.97 ± 0.40.7 ± 0.28380.17 ± 110.270.08 ± 0.072.77 ± 1.754.47 ± 3.88 p 0.1420.5810.5230.8040.7340.5720.583 SOD2 rs4880 AA421.37 ± 0.310.97 ± 0.450.75 ± 0.43439.12 ± 258.960.07 ± 0.063.53 ± 2.685.2 ± 3.68AG + GG831.37 ± 0.351.02 ± 0.50.79 ± 0.49380.54 ± 138.110.09 ± 0.112.72 ± 1.674.42 ± 4.25 p 0.9220.9370.8370.0300.7110.2540.089Data expressed in mean ± SD, Mann Whitney U test, p < 0.05 Comparison of SOD2 rs4880 and GPX3 rs3792797 genotypes in terms of oxidative stress variables Data expressed in mean ± SD, Mann Whitney U test, p < 0.05 Diagnostic performance of TrxR1 and PRDX1 as biomarkers for bipolar disorder The diagnostic performances of TrxR1 and PRDX1 as biomarkers were evaluated. These analyses were performed to differentiate the overall bipolar disorder group, regardless of mood phase, from healthy controls rather than to distinguish between bipolar phases (mania, depression, and euthymia). As shown in Table 5, the discrimination of TrxR1 and PRDX1 parameters in the diagnosis of bipolar disorder was statistically significant (p < 0.001, p < 0.001). The area under the ROC curve (AUC) for TrxR1 values to discriminate the diagnosis of bipolar disorder was 0.943. The sensitivity and specificity of TrxR1 values with a cut-off value ≥ 1.59 for differentiating bipolar disorder from healthy controls were 89% and 89%, respectively (Fig. 1). The area under the ROC curve (AUC) for PRDX1 values to discriminate the diagnosis of bipolar disorder was 0.943. The sensitivity and specificity of PRDX1 values with a cut-off value ≥ 1.73 for differentiating bipolar disorder from healthy controls were 87% and 86%, respectively (Fig. 2). Table 5Diagnostic performance of TrxR1 and PRDX1 in the diagnosis of bipolar disorderAUC95% CICut-offSensitivity (%)Specificity (%) p TrxR10.9430.878-1.00≥ 1.590.8960.893< 0.001PRDX10.9430.905–0.981≥ 1.730.8790.862< 0.001AUC: Area under the curve, CI: Confidence interval Diagnostic performance of TrxR1 and PRDX1 in the diagnosis of bipolar disorder AUC: Area under the curve, CI: Confidence interval Assessment of discriminative ability: ROC curve for TrxR1 Assessment of discriminative ability: ROC curve for PRDX1 Discussion In this study, bipolar disorder (BD) patients in different mood episodes and healthy controls were compared in terms of oxidative stress-related variables. Total antioxidant status (TAS), glutathione (GSH), thioredoxin reductase 1 (TrxR1), and peroxiredoxin 1 (PRDX1) levels were significantly different between patients with BD and healthy controls. Following Bonferroni correction for multiple testing, the differences in GSH, TrxR1, and PRDX1 remained statistically significant. No significant difference was found between the groups in terms of total oxidant status (TOS), superoxide dismutase (SOD) levels, and OSI score. Different mood episodes had no effect on TAS, GSH, TrxR1, and PRDX1 levels. The diagnostic performance of TrxR1 and PRDX1 as biomarkers in the diagnosis of BD was evaluated, and the diagnostic performance of these two parameters as biomarkers was found to be significant. In addition, participants were evaluated in terms of SOD2 rs4880 and GPX3 rs3792797 single nucleotide polymorphisms (SNPs). It was observed that there was a difference between patients with BD and healthy controls in terms of SOD2 rs4880. Participants with the AA allele had higher SOD levels than those with the AG and GG alleles. TAS levels were higher in patients with bipolar disorder than in healthy controls in the uncorrected analysis. However, this difference did not remain statistically significant after Bonferroni correction and should therefore be interpreted cautiously. The relationship between total antioxidant status (TAS), total oxidant status (TOS), and bipolar disorder (BD) has been the focus of many studies. Some studies have shown that patients with bipolar disorder have higher TAS, TOS, and OSI levels compared to healthy controls, indicating that oxidative balance is impaired [3, 21, 65]. On the other hand, in a study that included only patients in the manic episode, no difference was found in TOS levels, while TAS levels were found to be lower [42]. Additionally, in another study comparing manic and euthymic patients with healthy controls, no significant difference was observed between the groups [14]. In a recent study, as we found in our study, no significant difference was observed in TOS values between bipolar patients and healthy controls, while an increase was observed in TAS values [54]. TAS provides an estimation of extracellular antioxidants [57]. The increase in TAS may be interpreted as a metabolic response to suppress oxidant increase. On the other hand, TOS, which is an indicator of ROS, and OSI, which reflects oxidative stress levels, may be influenced by various parameters in patients with differing clinical characteristics; variations in sample sizes and study designs may also contribute to differing results. In most studies, only manic or euthymic patients were included in the bipolar disorder group, while very few studies have compared different mood episodes [3, 14, 21, 42, 65]. In these studies, consistent with our findings, no significant difference was observed in TAS and TOS levels between different mood episodes [14, 21]. The relationship between superoxide dismutase (SOD) and bipolar disorder (BD) has been investigated in various studies, and it has been shown that oxidative stress has a complex interaction with BD pathophysiology. In a study that compared depressive, manic, and euthymic bipolar patients with healthy controls, SOD values in the mania and depression groups were higher than those of healthy controls, whereas this difference was not observed in euthymic patients [40]. On the other hand, Selek et al., found that SOD activity was significantly lower in depressed bipolar patients [56]. In another study comparing euthymic and depressed bipolar patients with healthy controls, no difference was found between the groups [43]. In a meta-analysis comparing oxidative stress markers between BD and healthy controls, no difference in SOD was found between healthy controls and BD patients [32]. In our study, no significant difference was observed between bipolar patients in different mood episodes and healthy controls. Additionally, SOD values of bipolar patients using antidepressants and lamotrigine were found to be higher than those of other patients. Antidepressants and lamotrigine have been reported to increase SOD activity and expression in animal studies and cell culture models ([2, 41]; Xu, Steven Richardson, & Li, 2003). In a study by Selek et al., it was reported that SOD levels of bipolar patients, some of whom used antidepressants during the depressive episode, increased 30 days after treatment [56]. In another study that followed depressed patients, it was observed that SOD activities increased after 8 weeks of antidepressant treatment [30]. It has been suggested that antidepressants may have a neuroprotective effect by increasing SOD activity [41, 63]. A study conducted on adult epileptic patients treated with lamotrigine found increased SOD levels compared to the control group [50]. The use of antidepressants and lamotrigine, which was not taken into account in most studies examining the relationship between BD and SOD activity, may be among the factors explaining the different results observed in these studies. Many studies have shown that BD patients have lower GSH levels compared to healthy controls [26, 53, 55]. This lower level has been reported to indicate higher oxidative stress. Most studies have focused on bipolar patients in euthymic or manic episodes; studies comparing GSH levels of patients in different mood episodes are limited and often involve a low number of participants [22, 32, 51]. In a meta-analysis, it was found that GSH levels were significantly lower in BD patients compared to healthy controls, but there was no difference between manic and euthymic episodes [32]. This situation has been interpreted as suggesting that GSH could be considered a trait marker rather than a state marker capable of distinguishing different mood states. Our results are compatible with the literature and are important for demonstrating that there is no difference between GSH levels in different mood episodes. Taken together, these findings support the value of evaluating oxidative stress in bipolar disorder at multiple biological levels. Instead of focusing on a single biomarker, we assessed global oxidant-antioxidant balance, established antioxidant defense mechanisms, and the more specific thioredoxin/peroxiredoxin pathway within the same sample. This integrated approach was important because previous studies have reported inconsistent findings, possibly due in part to differences in the types of biomarkers examined. In our sample, the most prominent differences were observed in TrxR1 and PRDX1, suggesting that thiol-based redox pathways may be particularly relevant in bipolar disorder, whereas broader or more classical antioxidant markers did not show equally consistent alterations across groups. In line with this interpretation, the exploratory correlation analyses suggested that markers belonging to the same antioxidant pathway may be more closely related to each other than to global measures such as total antioxidant status. This pattern provides partial support for the integrated redox framework of the study and suggests that oxidative stress in bipolar disorder may involve related but not fully overlapping biological levels. However, these findings should be interpreted cautiously given the exploratory and cross-sectional nature of the analyses. Thioredoxin system components such as TrxR1 and the peroxiredoxin family, of which PRDX1 is a member, help reduce oxidative stress observed in Alzheimer’s, Parkinson’s, and Huntington’s diseases and play an important role in the pathogenesis of these diseases [9, 49, 58]. Studies have reported that TrxR1 and PRDX1 may be useful markers of oxidative stress in psychiatric disorders such as autism spectrum disorder and obsessive-compulsive disorder [5, 8, 17, 37, 67]. In a study on patients with irritable bowel syndrome, higher serum PRDX1 levels were found to be associated with increased anxiety and depression symptoms [67]. On the other hand, two studies in patients with major depressive disorder found that PRDX1 levels were lower [39, 48]. However, to our knowledge, no study investigating TrxR1 and PRDX1 levels in bipolar disorder has been conducted before. In this study, TrxR1 and PRDX1 levels were found to be significantly higher in patients with bipolar disorder. Furthermore, the diagnostic performance of these parameters was also found to be significant. However, the influence of different mood states on TrxR1 and PRDX1 levels has suggested that it cannot be considered a state marker. The diagnosis of psychiatric disorders, including bipolar disorder, is made through listening to an individual’s experiences and evaluating their behaviors. The heterogeneity within psychiatric disorders and the lack of definitive, objective biomarkers make accurate diagnosis and effective treatment in psychiatry highly challenging [33]. In bipolar disorder, hypomanic symptoms are often experienced as short periods with increased functionality, which can lead to these episodes being overlooked in diagnosis [60]. This can result in delays in diagnosing bipolar disorder. Due to challenges in early diagnosis, it is known that individuals with bipolar disorder may take approximately ten years to receive an accurate diagnosis [16, 31]. Consequently, in recent years, research on peripheral biomarkers in bipolar disorder has increased [11]. However, it can be stated that we are still far from identifying definitive and objective biomarkers [33]. TrxR1 and PRDX1 may be promising targets as trait biomarkers in bipolar disorder. In this study, the rates of SOD2 rs4880 AG and GG genotypes were found to be different in bipolar patients compared to healthy controls, while no difference was found between bipolar and healthy controls regarding GPX3 rs3792797 genotypes. However, both single SNPs have previously been associated with bipolar disorder [15, 25, 68]. The difference may be due to sample size and regional genetic differences. The SOD2 gene is located on chromosome 6q5, and the rs4880 SNP causes a substitution of A to G, resulting in the conversion of valine to alanine at position 16 [10]. Therefore, in our study, the AA homozygous group was compared with G allele carriers. A previous study conducted on schizophrenia patients found that G allele carriers exhibited weaker attention compared to the AA homozygous group [66]. We found that SOD levels were lower in individuals with SOD2 rs4880 AG and GG genotypes. The lower capacity to convert ROS into H₂O₂ indicates that the G allele has higher superoxide radical levels compared to the A allele. SOD2 rs4880 has been reported to be associated with pathologies such as diabetic nephropathy, diabetic retinopathy, coronary heart disease, and atherosclerosis [1, 6, 61]. Bipolar disorder is significantly associated with disability-adjusted life years, and part of this disability is linked to an increased risk of chronic systemic diseases, including cardiovascular diseases [13, 23]. In the literature, this comorbidity has been reported to potentially arise from shared genetic or biological alterations, such as interconnected immune-inflammatory and oxidative stress pathways [13, 32]. Although this study has many important results, it also has some limitations. A small sample size and the inclusion of chronic patients who are still undergoing treatment are among these limitations. These patients differ in terms of illness duration, the number of mood episodes experienced, the treatments they receive, and their medication doses. These factors may have influenced the assessed parameters. Although patients and controls live in the same city and have similar lifestyles, individual variations in lifestyle may exist. Factors such as dietary habits, and exercise routines can affect oxidative stress parameters. Moreover, the limited sample size did not allow for additional analyses controlling for potential confounding variables such as age, sex, body mass index, and medication use. In addition, no separate a priori power analysis was performed for the genetic component of the study. Therefore, the sample size may have been insufficient to detect modest genetic effects, and the SNP-related findings should be interpreted with caution. These potential influences, along with the cross-sectional design of the study, are also among its limitations. A study design with a larger sample size, better control of confounding factors, and long-term follow-up may more clearly explain the role of oxidative stress parameters in the etiopathogenesis of bipolar disorder. In this study, the examination of oxidative stress-related variables in bipolar disorder patients during different mood episodes provided important findings. This study represents an important step in understanding the role of oxidative stress in bipolar disorder. We think that the role of TrxR1 and PRDX1 in pathogenesis as a new biomarker may be significant. However, these findings need to be supported by larger samples and long-term studies. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements Not applicable. Abbreviations Bipolar disorder Reactive oxygen species Total antioxidant status Total oxidant status Oxidative stress index Superoxide dismutase Glutathione Single nucleotide polymorphisms Glutathione peroxidase 3 Thioredoxin reductase 1 Peroxiredoxin 1 Structured clinical interview for DSM-V clinical version Body mass index The hamilton depression rating scale-17 The young mania rating scale Author contributions Onur Gökçen: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review and editing. Kader Semra KARATAŞ: Formal analysis, Investigation, Methodology, Resources, Software, Supervision, Writing – review and editing. Merve AKKUŞ: Formal analysis, Investigation, Methodology, Resources, Software, Writing – review and editing. Feyza DÖNMEZ: Formal analysis, Investigation, Methodology, Resources, Software. Ayşe Koçak Sezgin: Data curation, Formal analysis, Methodology, Software, Supervision, Validation, Writing – original draft. Meliha Koldemir Gündüz: Data curation, Formal analysis, Methodology, Software, Supervision, Validation, Writing – original draft. Güllü Kaymak: Data curation, Formal analysis, Methodology, Software, Supervision, Validation, Writing – original draft. Funding The present study was financially supported by the Scientific Research Projects Coordination Unit of Kütahya Health Sciences University, Grant No: TSA-2024-172. Data availability The datasets used during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Ethical approval was obtained from the Independent Ethics Committee of Kütahya Health Sciences University (Decision No: 2023/06–19, dated May 11, 2023). Written informed consent was obtained from all participants. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets used during the current study are available from the corresponding author on reasonable request.