Xu XP, Ding S, Che YM, Li J, Zhang Z, Qian C. Expression and clinical value of the plasma peroxiredoxin family in early pregnancy for predicting subsequent gestational diabetes mellitus. World J Diabetes 2026; 17(8): 123635 [DOI: 10.4239/wjd.123635]
Corresponding Author of This Article
Chen Qian, Chief Technologist, Department of Clinical Laboratory, Central Laboratory, Xishan People’s Hospital of Wuxi City, No. 1128 Dacheng Road, Wuxi 214105, Jiangsu Province, China. qianchenwxxs@163.com
Research Domain of This Article
Obstetrics & Gynecology
Article-Type of This Article
research-article
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Xu XP, Ding S, Che YM, Li J, Zhang Z, Qian C. Expression and clinical value of the plasma peroxiredoxin family in early pregnancy for predicting subsequent gestational diabetes mellitus. World J Diabetes 2026; 17(8): 123635 [DOI: 10.4239/wjd.123635]
Xiao-Peng Xu, Zhou Zhang, Chen Qian, Department of Clinical Laboratory, Central Laboratory, Xishan People’s Hospital of Wuxi City, Wuxi 214105, Jiangsu Province, China
Shun Ding, Department of Clinical Laboratory, The 904th Hospital of Joint Logistic Support Force of PLA, Wuxi 214000, Jiangsu Province, China
Yan-Mei Che, Department of Obstetrics and Gynecology, Xishan People’s Hospital of Wuxi City, Wuxi 214105, Jiangsu Province, China
Jing Li, Department of Endocrinology, Jiangsu Province (Suqian) Hospital, Suqian 223800, Jiangsu Province, China
Co-corresponding authors: Zhou Zhang and Chen Qian.
Author contributions: Xu XP and Qian C were responsible for experiment conception and design; Ding S and Che YM carried out the experiments; Xu XP, Zhang Z, and Li J participated in data analysis; Xu XP, Zhang Z, Li J, and Qian C drafted the paper; Zhang Z and Qian C contributed equally as co-corresponding authors. All authors have read and approved the final manuscript to be published.
AI contribution statement: The authors take full responsibility and accountability for all content of this manuscript, including any portions for which AI tools were used as assistive technologies. All AI-assisted outputs were carefully reviewed, validated, and approved by the authors. AI tools were not used to generate original scientific data, perform independent scientific analyses, or draw scientific conclusions.
Supported by Scientific Research Program of Wuxi Health Commission, No. Z202515; and Youth Innovation Cultivation Program of Xishan People’s Hospital, No. 202104.
Institutional review board statement: This study obtained approval from the Ethics Committee of Xishan People’s Hospital of Wuxi City, No. xs2024ky050.
Informed consent statement: Verbal informed consent was obtained from all patients prior to the use of their stored samples and clinical data.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
STROBE statement: The authors have read the STROBE Statement-checklist of items, and the manuscript was prepared and revised according to the STROBE Statement-checklist of items.
Data sharing statement: All data can be obtained upon reasonable request from the corresponding author.
Corresponding author: Chen Qian, Chief Technologist, Department of Clinical Laboratory, Central Laboratory, Xishan People’s Hospital of Wuxi City, No. 1128 Dacheng Road, Wuxi 214105, Jiangsu Province, China. qianchenwxxs@163.com
Received: May 25, 2026 Revised: June 10, 2026 Accepted: July 2, 2026 Published online: August 15, 2026 Processing time: 73 Days and 16.6 Hours
Abstract
BACKGROUND
Many biomarkers have been developed to facilitate early gestational diabetes mellitus (GDM) diagnosis, but they have multiple limitations, including low accuracy and constraints in detection methods. The majority of biomarkers are influenced by gestational age, sample sizes and types, and analytical methodologies. Thus, there is a need of more accurate biomarkers for early diagnosis.
AIM
To investigate the expression of the peroxiredoxin (PRDX) family in early pregnancy and its clinical value for predicting subsequent GDM.
METHODS
Plasma samples and clinical data were collected at < 12 weeks of gestation from 222 women who subsequently developed GDM (confirmed by the 75 g oral glucose tolerance test at 24-28 weeks) and 222 women who remained normoglycemic throughout pregnancy, using random sampling at our hospital between January 2019 and December 2025. Participants were assigned to the observation group and control group accordingly. Plasma PRDX (PRDX1-PRDX6) levels were measured by ELISA. Receiver operating characteristic curves were applied for the assessment of the diagnostic value of PRDXs for early GDM. Univariate and multivariate logistic regression analyses were performed to identify related independent risk factors.
RESULTS
Baseline characteristics [age, pre-pregnancy body mass index (BMI), mid-pregnancy BMI] and biochemical indicators (triglycerides, low-density lipoprotein-cholesterol, fasting blood glucose, glycated hemoglobin) were markedly higher in the GDM group (P < 0.05). Expression levels of all PRDX family members in plasma showed obvious elevations in the observation group (P < 0.05). PRDX1, PRDX4, and PRDX5 displayed high diagnostic value for early GDM, with areas under the curve of 0.757, 0.776, and 0.742, respectively, and a combined diagnostic area under the curve of 0.836. Univariate analysis identified 15 significant risk factors, and multivariate analysis revealed age, pre-pregnancy BMI, triglycerides, low-density lipoprotein-cholesterol, fasting blood glucose, glycated hemoglobin, PRDX1, and PRDX3 as independent risk factors for early GDM (P < 0.05).
CONCLUSION
All PRDX family members showed significantly elevated expression in early GDM. PRDX1, PRDX4, and PRDX5 exhibited substantial diagnostic value, while PRDX1 and PRDX3 were identified as independent risk factors.
Core Tip: This study provides the first comprehensive assessment of the plasma peroxiredoxin (PRDX) family in early pregnancy as predictive biomarkers for gestational diabetes mellitus. PRDX1, PRDX4, and PRDX5 demonstrated strong individual diagnostic performance (areas under the curve: 0.757, 0.776, and 0.742, respectively), and their combination further improved accuracy to 0.836. Furthermore, PRDX1 and PRDX3 were identified as independent risk factors for early gestational diabetes mellitus, highlighting the potential of the PRDX family for early screening and risk stratification.
Citation: Xu XP, Ding S, Che YM, Li J, Zhang Z, Qian C. Expression and clinical value of the plasma peroxiredoxin family in early pregnancy for predicting subsequent gestational diabetes mellitus. World J Diabetes 2026; 17(8): 123635
Gestational diabetes mellitus (GDM) refers to varying degrees of impaired glucose tolerance first occurring or detected during pregnancy[1]. GDM is related to obstetric and neonatal complications primarily due to increased birth weight, and hence is considered a risk factor for future cardiometabolic diseases in both mothers and offspring[2,3]. With the gradual improvement in living conditions and changing fertility concepts, the prevalence of GDM has increased. This increase is attributed to the rising body mass index (BMI) among women of reproductive age, inadequate control of blood glucose and lipid levels, and advanced maternal age[1,4]. Currently, GDM is diagnosed primarily by the oral glucose tolerance test (OGTT). However, this method has limitations, including issues related to user experience, measurement accuracy, controversies in diagnostic criteria, and applicability in specific populations. Early diagnosis of GDM is essential, as pregnancy outcomes are better in women diagnosed early than those diagnosed later[5,6]. Many biomarkers have been developed to facilitate early GDM diagnosis, but they also face multiple limitations, including low accuracy and constraints in detection methods[7]. The majority of biomarkers are influenced by gestational age, sample sizes and types, and analytical methodologies[8]. Thus, there is a need of more accurate biomarkers for early diagnosis. Prognostic monitoring in GDM also encounters challenges, including poor compliance, insufficient predictive accuracy, and gaps between monitoring and intervention. Therefore, the identification of a highly sensitive and specific indicator with prognostic monitoring potential for GDM patients is urgently required.
Peroxiredoxin (PRDX) belongs to a highly conserved peroxidase family consisting of six members (PRDX1-PRDX6). This enzyme family regulates downstream signaling by maintaining a dynamic equilibrium between hydrogen peroxide and degrading peroxynitrite and lipid peroxides under conditions of oxidative stress. Consequently, PRDX modulates intracellular peroxide levels and affects gene expression, cell proliferation, apoptosis, migration, differentiation, and other biological processes. Studies have indicated that PRDX3 positively contributes to insulin secretion in a glucose concentration-dependent manner, making it a potential marker for high insulin resistance[9]. Additionally, research has demonstrated downregulated PRDX6 expression in the adipose tissue of women with GDM, assisting in explaining the role of visceral obesity in GDM pathogenesis from new perspectives[10].
This study aims to investigate plasma expression of the PRDX family members in pregnant women with early GDM. Furthermore, the diagnostic and prognostic monitoring value of these biomarkers for early GDM is evaluated, thereby providing new targets for disease diagnosis and treatment in clinical practice.
MATERIALS AND METHODS
Study subjects
This observational study included pregnant women who attended their first prenatal visit at < 12 weeks of gestation at our hospital between January 2019 and December 2025. Peripheral blood samples and clinical data were collected at this initial visit. All participants were followed up, and a 75 g OGTT was performed at 24-28 weeks of gestation. Based on the OGTT results, women were classified into the observation group (n = 222, those diagnosed with GDM) and the control group (n = 222, those who remained normoglycemic throughout pregnancy). The two groups were randomly selected from eligible participants using a random number table. GDM was diagnosed according to clinical criteria using a single-threshold approach[11]. The diagnosis of GDM should meet one or more of the following criteria: (1) Fasting plasma glucose ≥ 5.1 mmol/L; (2) 1-hour OGTT plasma glucose ≥ 10.0 mmol/L; or (3) 2-hour OGTT plasma glucose ≥ 8.5 mmol/L. Inclusion criteria were as follows: (1) Singleton pregnancy; (2) First prenatal visit and completion of fasting plasma glucose and glycated hemoglobin (HbA1c) tests at < 12 weeks of gestation; and (3) Complete clinical records, including detailed pregnancy and maternity history, physical examination, and laboratory test results. Exclusion criteria were: (1) Diagnosis of type 1 diabetes mellitus or type 2 diabetes mellitus (T2DM) before pregnancy; (2) Presence of major organ dysfunction; (3) Use of medications affecting glucose metabolism; (4) Women with threatened abortion or pregnancy conceived through assisted reproductive technology; and (5) Pregnancies complicated by severe fetal malformations or chromosomal abnormalities. In this study, “early GDM” refers to pregnant women who were normoglycemic at < 12 weeks of gestation but were subsequently diagnosed with GDM following the 75 g OGTT at 24-28 weeks. This study acquired approval from the Ethics Committee of Xishan People’s Hospital, No. xs2024ky050. Verbal informed consent was obtained from all patients prior to the use of their stored samples and clinical data.
Biochemical marker testing
All biochemical markers, including triglycerides (TG), total cholesterol (TC), high-density lipoprotein-cholesterol (HDL-C), low-density lipoprotein-cholesterol (LDL-C), fasting blood glucose (FBG), and HbA1c, were measured by the Laboratory Department of Xishan People’s Hospital. TG, TC, HDL-C, LDL-C, and FBG were tested using a LABOSPECT 008 AS fully automated biochemical analyzer and corresponding reagents [Hitachi High-tech (Shanghai) International Trade Co., Ltd., Shanghai, China]. HbA1c was analyzed using an HA180 HbA1c analyzer (Tianjin Aike Medical Equipment Co., Ltd., Tianjin, China).
Measurement of PRDX family expression levels
Expression levels of each PRDX family member were determined using ELISA kits: Human PRDX1 ELISA Kit (JL19445), Human PRDX2 ELISA Kit (JL45658), Human PRDX3 ELISA Kit (JL15455), Human PRDX5 ELISA Kit (JL12625) and Human PRDX6 ELISA Kit (JL48979) were obtained from Shanghai Jianglai Biotechnology, and Human PRDX4 ELISA Kit (CSB-E13442h) from Wuhan Huamei Biotech.
Initially, the required number of microplate strips was isolated from the aluminum foil pouch to undergo 20 minutes of equilibration at room temperature. Wells were designated for standards and samples, with the standard wells containing 50 μL of each standard solution and the sample wells containing 10 μL of the sample and 40 μL of sample diluent. Except for blank wells, each well was supplemented with 100 μL of HRP-labeled detecting antibody. The plates were sealed and incubated for 1 hour at 37 °C. After incubation, liquid was aspirated from the wells, and the plates were blotted dry with absorbent paper. Each well was filled with wash buffer and left for 1 minutes, the buffer was then discarded and the plate blotted dry. These steps were repeated five times. Each well was subsequently supplemented with 50 μL of substrate solution A and 50 μL of substrate solution B. The plates were then incubated for 15 minutes at 37 °C in the dark. Then, 50 μL of stop solution was added to each well, and a microplate reader was used to measure optical density at 450 nm within 15 minutes.
Data and sample collection timeline
Peripheral blood samples and clinical data (including age, height, pre-pregnancy weight, and birth history) were collected at the first prenatal visit (gestational age < 12 weeks). Pre-pregnancy weight and BMI were self-reported and verified at this visit. Mid-pregnancy weight and BMI were obtained from medical records at 24-28 weeks of gestation during the OGTT visit. All biochemical markers (TG, TC, HDL-C, LDL-C, FBG, HbA1c, PRDX family) were measured from the early pregnancy blood sample.
Statistical analysis
Statistical analysis was conducted using SPSS version 26.0. The Kolmogorov-Smirnov test was employed for the assessment of normality of continuous data, and Levene’s test evaluated homogeneity of variance. Normally distributed data with homogenous variance are presented as mean ± SD and the independent samples t test was conducted for comparison. Data not meeting these conditions are presented as median (25% quartile-75% quartile) and the Mann-Whitney U test was used for relevant comparison. Count data are presented as n (%) and the χ² test was used for relevant comparison. Receiver operating characteristic curves were constructed for diagnostic performance evaluation of plasma PRDXs for early GDM, with the area under the curve (AUC) indicating diagnostic accuracy. A higher AUC approaching 1.0 represented superior diagnostic performance, and an AUC > 0.7 was considered good. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors for early GDM. A P value < 0.05 was considered statistically significant. Prior to multivariate logistic regression, collinearity among candidate variables was assessed using the variance inflation factor (VIF). Variables with VIF ≥ 5 indicated significant collinearity. Pre-pregnancy weight, mid-pregnancy weight, and mid-pregnancy BMI showed VIF values exceeding 5 when entered together with pre-pregnancy BMI. Therefore, to avoid unstable estimates due to multicollinearity, only pre-pregnancy BMI was retained in the final multivariate analysis as the most clinically standard indicator of maternal adiposity.
RESULTS
Normality and homogeneity of variance assessment of plasma indicators between the groups
Initially, the normality and variance homogeneity of all quantitative data were evaluated. The results indicated that quantitative plasma indicators in the control and observation groups did not simultaneously meet the assumptions of normality and homogeneity of variance. Therefore, the study adopted the Mann-Whitney U test for the statistical analysis of indicators including baseline parameters, TG, TC, HDL-C, LDL-C, FBG, HbA1c, PRDX1, PRDX2, PRDX3, PRDX4, PRDX5, and PRDX6 (Table 1).
Table 1 Comparison of normality and homogeneity of variance for plasma indicators between the control and observation group.
Comparison of baseline characteristics between the groups
Baseline characteristics, including age, height, pre-pregnancy weight, pre-pregnancy BMI, mid-pregnancy weight, mid-pregnancy BMI, and birth history, were compared. Age, pre-pregnancy weight, pre-pregnancy BMI, mid-pregnancy weight, and mid-pregnancy BMI were significantly higher in the observation group (P < 0.05; Table 2).
Table 2 Comparison of baseline characteristics between the control and observation group.
Comparison of plasma biochemical parameters between the groups
Plasma biochemical indicators, including TG, TC, HDL-C, LDL-C, FBG, and HbA1c, were compared. The observation group demonstrated obviously higher levels of TGs, LDL-C, FBG, and HbA1c, but markedly lower HDL-C levels (P < 0.05; Table 3).
Table 3 Comparison of plasma biochemical parameters between the control and observation group.
Comparison of plasma PRDX family member expression between the groups
Plasma expression levels of PRDX family members were compared. Their expression levels increased to varying degrees in the observation group (P < 0.05; Table 4).
Table 4 Comparison of plasma peroxiredoxin family member expression between the control and observation group.
Diagnostic value of plasma PRDX family members in women with early GDM
Receiver operating characteristic curves were constructed to measure the diagnostic value of plasma PRDX family members in detecting early GDM. According to the results, PRDX1, PRDX4, and PRDX5 had strong diagnostic performance, with AUC values of 0.757, 0.776, and 0.742, respectively, all surpassing 0.7. Their optimal cutoff values were 52.80, 123.10, and 65.51, with corresponding sensitivities and specificities of (0.874, 0.550), (0.644, 0.788), and (0.946, 0.514), respectively, demonstrating high diagnostic potential for early GDM. Subsequently, the combination of PRDX1, PRDX4, and PRDX5 was assessed, yielding an enhanced AUC of 0.836 and corresponding sensitivity and specificity of (0.851, 0.703) (Figure 1 and Table 5). Thus, PRDX1, PRDX4, and PRDX5 showed significant individual diagnostic potential, with their combined analysis further improving diagnostic accuracy.
Figure 1 Diagnostic value of plasma peroxiredoxin family members in women with early gestational diabetes mellitus.
Combined: Indicates the joint receiver operating characteristic analysis of peroxiredoxin 1 (PRDX1), PRDX4 and PRDX5. PRDX: Peroxiredoxin.
Table 5 Diagnostic value of plasma peroxiredoxin family members in women with early gestational diabetes mellitus.
Univariate logistic regression analysis of early GDM
Univariate logistic regression analysis was performed to identify potential risk factors for early GDM. The analysis revealed that older age, higher pre-pregnancy weight, higher pre-pregnancy BMI, greater mid-pregnancy weight, greater mid-pregnancy BMI, elevated TG, higher LDL-C, increased FBG, and higher HbA1c were all significant risk factors (all P < 0.001), whereas HDL-C acted as a significant protective factor [odds ratio (OR) = 0.554, P = 0.039]. Notably, all six PRDX family members showed significant positive associations with early GDM (all P < 0.001). In total, 15 variables were identified as significant risk factors (Table 6). These findings provided the basis for subsequent multivariate analysis.
Table 6 Univariate logistic regression analysis in women with early gestational diabetes mellitus.
Multivariate logistic regression analysis of early GDM
To avoid multicollinearity among highly correlated anthropometric variables, pre-pregnancy weight, mid-pregnancy weight, and mid-pregnancy BMI were excluded from the multivariate analysis, and only pre-pregnancy BMI was retained as the representative adiposity indicator. All significant univariate factors were then entered into multivariate logistic regression. After adjustment, the following variables remained independent risk factors for early GDM: Age (OR = 1.299, P < 0.001), pre-pregnancy BMI (OR = 1.340, P < 0.001), TG (OR = 1.923, P < 0.001), LDL-C (OR = 20.624, P < 0.001), FBG (OR = 4.105, P = 0.001), HbA1c (OR = 5.532, P < 0.001), PRDX1 (OR = 1.012, P = 0.009), and PRDX3 (OR = 1.007, P = 0.015). HDL-C lost statistical significance after adjustment (P = 0.154). Among the PRDX family, only PRDX1 and PRDX3 remained independent predictors, whereas PRDX2, PRDX4, PRDX5, and PRDX6 were no longer significant (all P > 0.05). These results are summarized in Table 7.
Table 7 Multivariate logistic regression analysis in women with early gestational diabetes mellitus.
GDM substantially threatens maternal and fetal health. In pregnant women, GDM considerably elevates the incidence of birth injury, cesarean section, gestational hypertension, polyhydramnios, and macrosomia. Additionally, it elevates their long-term risk of developing T2DM[12,13]. In the fetus and newborn, GDM can cause abnormal intrauterine growth, preterm delivery, neonatal respiratory distress syndrome, hypoglycemia, and may impact future metabolic health, increasing susceptibility to childhood obesity and diabetes mellitus[14,15]. Therefore, early screening and standardized management of early GDM are essential.
Currently, GDM screening mainly involves conducting an OGTT between 24 and 28 weeks of gestation[16], preceded by risk factor assessment by clinicians. Early and intensive screening strategies are adopted for high-risk groups, including those with advanced maternal age, overweight or obesity, history of GDM, family history of diabetes mellitus, elevated FBG, or increased HbA1c levels at the initial prenatal visit[17]. If initial screening is abnormal, a repeat OGTT may be performed in the second trimester to confirm the diagnosis. However, current risk stratification has limitations[18], mainly due to its dependence on static and general clinical parameters, leading to inadequate sensitivity[19]. Consequently, many pregnant women without typical risk factors remain at substantial risk of missed GDM diagnosis. Furthermore, this approach overlooks critical information such as gestational weight gain, dynamic metabolic changes, and regional variations[20]. Thus, identifying more precise biomarkers and determining independent risk factors for early GDM is crucial[21].
This study found increased baseline parameters in patients with early GDM. This suggests that these variables may significantly benefit the development of early GDM. Sun et al[22] also reported a strong correlation between pre-pregnancy overweight and obesity with GDM risk, which aligns with our findings. In terms of plasma biochemical markers, this study revealed elevated levels of TG, LDL-C, FBG, and HbA1c, alongside decreased HDL-C, in women with early GDM. Elevated blood lipids, fasting glucose, and HbA1c were associated with increased prevalence of early GDM, consistent with established knowledge. Two other studies similarly demonstrated that women with GDM presented higher HbA1c, FBG, LDL-C, and TG vs their counterparts, whereas HDL-C was comparatively lower[23,24].
As the core indicator of this study, PRDX family members have primarily been studied in cancer, neurodegenerative disorders, cardiovascular diseases, and metabolic syndromes, displaying considerable variation across different diseases and family members[25-29]. However, limited clinical research exists regarding their roles in endocrine disorders, particularly diabetes mellitus. Among PRDX family members, PRDX4 is most closely associated with diabetes mellitus and related endocrine and metabolic disorders, followed by PRDX3 and PRDX6. These three members play crucial roles in distinct pathological mechanisms underlying diabetes mellitus. Studies have shown that PRDX4 overexpression in glucose-stimulated INS-1E cells effectively metabolized luminal hydrogen peroxide, enhancing glucose-induced insulin secretion, proinsulin mRNA transcription, and insulin content[30]. PRDX4 also provides protective regulatory effects on diabetic retinopathy progression by destabilizing dipeptidyl peptidase-4 protein, suggesting that increased PRDX4 levels might represent a promising therapeutic strategy for diabetic retinopathy[31]. PRDX3 acetylation is mediated by impaired antioxidant defense. Moreover, teneligliptin ameliorates mitochondrial dysfunction by activating the SIRT1-mediated SIRT3-PRDX3 signaling pathway[32]. The S-palmitoylation of PRDX6 at Cys47 affects its interaction with the AE3 C-terminal domain, which regulates AE3 activity in the nervous system and prevents diabetic neuropathy[33]. Clearly, most studies that elucidate the linkage between PRDX family members and diabetes-related diseases have focused on fundamental mechanisms rather than clinical applications.
In the present study, plasma levels of all PRDX family members were elevated in women with early GDM. PRDX1, PRDX4, and PRDX5 exhibited high diagnostic value for early GDM, with a combined diagnostic accuracy (AUC) of 0.836. Thus, PRDX1, PRDX4, and PRDX5 may serve as reliable diagnostic biomarkers for early GDM in the future. This study also identified age, pre-pregnancy weight, pre-pregnancy BMI, mid-pregnancy weight, mid-pregnancy BMI, TG, LDL-C, FBG, HbA1c, and all PRDX family members as significant risk factors for early GDM. Further analyses confirmed age, pre-pregnancy BMI, TG, LDL-C, FBG, HbA1c, PRDX1, and PRDX3 as independent risk factors.
This study has several limitations. First, the single-center, retrospective case-control design limits generalizability, and the findings require external validation in independent, prospectively recruited cohorts. Second, as only plasma samples were stored in this retrospective study and peripheral blood cells or placental tissue were not collected, we were unable to validate the expression of the PRDX family at the mRNA level using quantitative reverse transcription-polymerase chain reaction or at the intracellular protein level using western blotting. This ruled out determination of whether the elevated plasma PRDX levels were due to increased cell expression, active secretion, or passive release. Although plasma ELISA is still a clinically relevant method for circulating secreted proteins, and our results reliably reflect the concentration of early pregnancy PRDX in the circulation, ELISA may be affected by antibody cross reactivity and cannot distinguish between reduced and oxidized forms of PRDX. Third, although we identified associations between PRDXs and subsequent GDM, causality cannot be inferred; the observed elevations may represent early compensatory responses to metabolic stress rather than pathogenic drivers. This interpretation is further limited by the absence of direct oxidative stress marker measurements, as the stored plasma was not pre-treated with antioxidant stabilizers and long-term storage rendered retrospective measurement of these labile markers unreliable. Fourth, potential confounders such as blood pressure, smoking, alcohol consumption, dietary patterns, physical activity, gestational weight gain trajectory, and family history of diabetes were not fully adjusted for, as these variables were not systematically documented in the medical records of this retrospective cohort. Fifth, the predictive model was not externally validated, and the reported AUCs may represent overfitted estimates. Future multicenter prospective studies with larger sample sizes, standardized collection of both plasma (with appropriate stabilizers) and peripheral blood mononuclear cells, metabolomic or proteomic validation, longitudinal post-GDM follow-up, and the application of flexible modeling strategies such as restricted cubic spline analysis are needed to fully characterize potential non-linear dose-response relationships and to confirm the clinical utility of the PRDX family in GDM screening and risk stratification.
CONCLUSION
This study demonstrated elevated plasma PRDX family member expression in women with early GDM. PRDX1, PRDX4, and PRDX5 displayed strong diagnostic performance, while PRDX1 and PRDX3 could independently predict the risk of early GDM, which highlights the clinical importance of PRDX family members, potentially opening new avenues for the clinical diagnosis and management of early GDM.
