Li SY, Hu MM, Xiang LL, Zhu YT, Liu Y, Chen YT, Chen YJ, Zeng Y, Zhong TY. Comparison of triglyceride-glucose index and its derived indices as markers for adverse pregnancy outcomes. World J Diabetes 2026; 17(5): 117022 [DOI: 10.4239/wjd.v17.i5.117022]
Corresponding Author of This Article
Tian-Ying Zhong, Professor, Department of Clinical Laboratory, Women’s Hospital of Nanjing Medical University, Nanjing Women and Children’s Healthcare Hospital, No. 123 Tianfei Alley, Mochou Road, Nanjing 210002, Jiangsu Province, China. zhongtianying@njmu.edu.cn
Research Domain of This Article
Endocrinology & Metabolism
Article-Type of This Article
Retrospective Cohort Study
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/
May 15, 2026 (publication date) through May 14, 2026
Times Cited of This Article
Times Cited (0)
Journal Information of This Article
Publication Name
World Journal of Diabetes
ISSN
1948-9358
Publisher of This Article
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Li SY, Hu MM, Xiang LL, Zhu YT, Liu Y, Chen YT, Chen YJ, Zeng Y, Zhong TY. Comparison of triglyceride-glucose index and its derived indices as markers for adverse pregnancy outcomes. World J Diabetes 2026; 17(5): 117022 [DOI: 10.4239/wjd.v17.i5.117022]
Co-corresponding authors: Yu Zeng and Tian-Ying Zhong.
Author contributions: Li SY and Hu MM designed the research plan, analyzed the data, participated in discussions, drafted the initial manuscript, and reviewed the manuscript; Xiang LL, Zhu YT and Liu Y analyzed the data, contributed to data presentation, and participated in revising the manuscript; Chen YT and Chen YJ participated in the discussion and provided further editing and comments; Li SY and Hu MM contributed equally to this work as co-first authors. Zeng Y and Zhong TY critically revised the manuscript and approved the final version. They contributed equally as co-corresponding authors. All authors approved the final version. Zeng Y and Zhong TY contributed equally to this work, due to their distinct yet complementary contributions essential to this multidisciplinary study. Zhong TY served as the principal investigator of the clinical cohort. She was responsible for the study design, the ethical approval process, and the rigorous oversight of clinical data collection, ensuring the integrity of the pregnancy outcome records. On the other hand, Zeng Y led the statistical analysis and the metabolic interpretation of the study. She was responsible for the calculation of the triglyceride-glucose derived indices, the advanced regression modeling, and the drafting of the manuscript's discussion regarding lipid metabolism mechanisms. Given that this research bridges clinical obstetrics with metabolic data science, the specific expertise of both authors was equally critical. Neither author could have completed the project independently without the other’s specialized leadership. Therefore, we believe designating them as co-corresponding authors best reflects their equal intellectual input and responsibility for this work.
Supported by Nanjing Medical Science and Technique Development Foundation, No. YKK23151; the Opening Foundation of Key Laboratory, No. JSHD202313; Yingke Xinchuang Research Foundation of Jiangsu Blood Transfusion Association, No. JSYK2024006; and Open Project of the State Key Laboratory of Reproductive Medicine of Nanjing Medical University, No. SKLRM-K202107.
Institutional review board statement: The study was reviewed and approved by the Medical Ethics Committee of Women’s Hospital of Nanjing Medical University, Nanjing Women and Children’s Healthcare Hospital (No. 2022KY-123).
Informed consent statement: Considering that this study was retrospective and that the data did not contain any personally identifiable data, the Medical Ethics Committee of Nanjing Women and Children’s Healthcare Hospital determined that the requirement for informed consent from participants was not applicable.
Conflict-of-interest statement: The authors declare that they have no conflicts of interest. Furthermore, there are no conflicts of interest between the funder and the authors.
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: The data that support the findings of this study are available from the corresponding author Yu Zeng upon reasonable request.
Corresponding author: Tian-Ying Zhong, Professor, Department of Clinical Laboratory, Women’s Hospital of Nanjing Medical University, Nanjing Women and Children’s Healthcare Hospital, No. 123 Tianfei Alley, Mochou Road, Nanjing 210002, Jiangsu Province, China. zhongtianying@njmu.edu.cn
Received: November 26, 2025 Revised: January 25, 2026 Accepted: March 16, 2026 Published online: May 15, 2026 Processing time: 165 Days and 20.7 Hours
Abstract
BACKGROUND
Insulin resistance (IR) and obesity are key risk factors for adverse pregnancy outcomes (APOs). While the triglyceride-glucose (TyG) index is a reliable surrogate for IR, its independent predictive utility during pregnancy may be limited. Consequently, novel composite indices combining the TyG index with obesity metrics have been proposed. However, a systematic comparison of their predictive performance for APOs is currently lacking.
AIM
To compare the predictive performance of the TyG index and its derived indices for APOs.
METHODS
A retrospective cohort study included 10422 pregnant women from Eastern China. The TyG index was calculated using fasting triglycerides and glucose. The TyG-derived indices were computed by integrating TyG with body mass index (BMI), waist circumference and waist-to-height ratio (WHtR). Logistic regression, restricted cubic spline (RCS) models, and receiver operating characteristic (ROC) analyses were utilized to evaluate associations and predictive performance. Subgroup analyses stratified by maternal age and pre-pregnancy BMI were conducted to assess consistency.
RESULTS
Higher levels of TyG and its derivatives were independently associated with increased risks of gestational diabetes mellitus (GDM), hypertensive disorders of pregnancy (HDP), cesarean delivery (CD), large for gestational age (LGA) (all P for trend < 0.05). After adjustment, TyG-BMI showed the strongest association with HDP [odds ratio (OR) = 1.46, 95% confidence interval (CI): 1.14-1.87] and LGA (OR = 3.28, 95%CI: 2.75-3.91) when comparing the top vs bottom tertile. RCS analysis revealed primarily linear associations, with non-linear trends observed for LGA. ROC analysis indicated that all derived indices demonstrated superior performance over the TyG index alone in predicting HDP, CD, and LGA, while TyG-WHtR provided a marginal improvement in GDM prediction. The robustness of the findings was validated through additional subgroup analyses.
CONCLUSION
Both TyG and its derived indices are valuable markers for the early prediction of APOs. Among these, TyG-BMI may represent a promising indicator for identifying the high-risk population.
Core Tip: This study demonstrates that integrating obesity metrics with the triglyceride-glucose (TyG) index significantly enhances the prediction of adverse pregnancy outcomes. The derived indices, particularly TyG-body mass index (TyG-BMI), showed superior predictive performance for hypertensive disorders of pregnancy, cesarean delivery, and large-for-gestational-age infants compared to the TyG index alone. These findings highlight that simple, cost-effective composite indices like TyG-BMI can serve as useful tools for early identification of high-risk pregnancies, offering significant clinical value.
Citation: Li SY, Hu MM, Xiang LL, Zhu YT, Liu Y, Chen YT, Chen YJ, Zeng Y, Zhong TY. Comparison of triglyceride-glucose index and its derived indices as markers for adverse pregnancy outcomes. World J Diabetes 2026; 17(5): 117022
Gestational diabetes mellitus (GDM) is one of the most prevalent complications during pregnancy[1], with a global prevalence of approximately 14%, exhibiting a continuous upward trend[2]. GDM significantly increases the risk of other adverse pregnancy outcomes (APO), including hypertensive disorders of pregnancy (HDP), cesarean delivery (CD), large for gestational age (LGA), placental abruption, premature rupture of membranes, preterm delivery, macrosomia, and postpartum hemorrhage, among others[3-6]. These complications pose risks to maternal and fetal health across both immediate and extended durations[7,8]. Therefore, early identification of clinical hazards is vital for implementing effective prevention and intervention measures for mothers and infants.
Physiological insulin resistance (IR) during pregnancy is an important adaptive change that ensures an adequate nutrient supply to the fetus. However, pathological IR can serve as a critical pathological basis for APOs, potentially affecting both maternal and fetal health. The hyperinsulinemic-euglycemic clamp (HEC) is the gold standard for IR testing but is impractical in clinics due to its complexity and cost[9]. The homeostasis model assessment of IR (HOMA-IR) is a common surrogate but requires precise insulin measurements. Research has shown the triglyceride-glucose (TyG) index is closely correlated with HEC test outcomes and HOMA-IR scores[10,11]. This index has proven to be a simple, economical, and clinically practical biomarker for IR assessment[12], demonstrating significant associations with various metabolic disorders, including diabetes mellitus[13], cardiovascular diseases[14], and hypertension[15]. The TyG index is increasingly utilized in obstetric populations[16-19]. Therefore, given the unique physiological state of IR during pregnancy, the specificity and predictive performance of TyG in this population require further refinement[16].
Furthermore, novel indices that combine the TyG index with obesity parameters, including TyG-body mass index (TyG-BMI), TyG-waist circumference (TyG-WC), and TyG-waist-to-height ratio (TyG-WHtR) have been proposed for assessing metabolic disease risk. TyG-BMI serves as a novel predictor of metabolic dysfunction-associated steatotic liver disease (formerly non-alcoholic fatty liver disease) in patients with type 2 diabetes mellitus[20]; TyG-WC predicts hypertension incidence in middle-aged and elderly populations[21]; and TyG-WHtR shows a strong correlation with cardiovascular disease incidence[22]. Crucially, the TyG index combined with BMI, WC, and WHtR can enhance hypertension risk stratification in individuals aged ≥ 45 years[21] and exhibits a considerable association with the incidence of atherosclerotic cardiovascular disease[23]. However, limited evidence exists regarding the link between the TyG index and its derived indices and APOs.
Therefore, based on a large population cohort, this study aimed to investigate the associations between TyG, TyG-BMI, TyG-WC, and TyG-WHtR during the second trimester and the risk of APOs. Furthermore, we aimed to elucidate and compare the predictive capabilities of these TyG-related indices for risks of APOs, providing novel insights for prevention and clinical management in obstetric practice.
MATERIALS AND METHODS
Participants
This retrospective observational cohort study analyzed data from women who underwent routine antenatal examinations at Nanjing Women and Children’s Healthcare Hospital between January 2021 and December 2022. The inclusion criteria included: (1) Participants aged 18-49 years; and (2) Singleton live birth. The exclusion criteria included: (1) Absence of 75-gram oral glucose tolerance test (OGTT) data; (2) Incomplete laboratory data; (3) Pre-existing severe acute or chronic diseases, including diabetes mellitus before pregnancy, lipid disorders requiring medication (familial or acquired), hepatic, renal, cardiovascular, autoimmune, and severe psychiatric conditions; and (4) Missing data for key variables. Ultimately, 10422 women with singleton deliveries were eligible for the final analysis. This research received ethical approval from the Ethics Committee of Nanjing Women and Children’s Healthcare Hospital (Approval No. 2022KY-123) and adhered to the Declaration of Helsinki. Given the retrospective design, the requirement for informed consent was waived, and all potentially identifiable patient information was removed.
Data collection and metrics
During the second trimester of pregnancy (around 24 weeks), anthropometric and biochemical evaluations were conducted during prenatal examinations. Data on the following were collected: Maternal age, pre-pregnancy BMI, WC, gravidity, parity, systolic blood pressure (SBP), and diastolic blood pressure (DBP). Following an 8-hour fast, blood samples were drawn to measure various biomarkers. These included alanine aminotransferase (ALT), aspartate aminotransferase (AST), total protein (TP), albumin (ALB), urea (UREA), creatinine (CREA), uric acid (UA), triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and glycated hemoglobin (HbA1c). OGTT outcomes measured at 24-28 gestational weeks were recorded. The specific calculation formulas for each index are listed here: BMI = body weight (kg)/height2 (m2). WHtR = WC (cm)/height (cm). TyG = Ln [TG (mg/dL) × fasting glucose (mg/dL)/2]. TyG-BMI = TyG × BMI. TyG-WC = TyG × WC. TyG-WHtR = TyG × WHtR.
Definition of APOs
Pregnancy complications were extracted from digital health records. Primary outcomes included: (1) GDM was identified if any glucose level following a 75-gram OGTT crossed the threshold: Fasting blood glucose at or above 5.1 mmol/L, 1-hour post-load glucose of 10.0 mmol/L or higher, or 2-hour post-load glucose levels hitting 8.5 mmol/L or more, based on the guidelines set by the International Association of Diabetes and Pregnancy Study Groups[24]; (2) HDP, comprising gestational hypertension, preeclampsia, chronic hypertension, and superimposed preeclampsia, were defined by the guidelines of the American College of Obstetricians and Gynecologists[25]; and (3) LGA was characterized as a birth weight surpassing the 90th percentile after adjustment for sex and gestational age, according to the Chinese neonatal birth weight standard curve[26]. Secondary outcomes encompassed gestational anemia, oligohydramnios (less than 300 mL), polyhydramnios (more than 2000 mL); postpartum anemia, preterm birth, fetal distress (non-reassuring heart rate or meconium staining), small for gestational age (SGA), fetal growth restriction, neonatal intensive care unit admission, and low Apgar score (< 7 at 1 or 5 minutes)[4].
Covariates
Potential confounders included: Maternal age, pre-pregnancy BMI, gravidity (1 or ≥ 2), parity (primiparous or multiparous), SBP, and DBP. We also accounted for biochemical confounders: ALT, AST, TP, ALB, UREA, CREA, UA, HDL-C, LDL-C and HbA1c. Due to the high collinearity between TC and LDL-C, only LDL-C was retained in the models. Pre-pregnancy BMI was excluded from models involving TyG-BMI due to its presence in the index calculation.
Statistical analysis
The study cohort of pregnant women was categorized into tertiles (T1-T3) determined by their values for the TyG index and its derivatives. The normality of continuous variables was assessed using the Kolmogorov-Smirnov test. Subsequently, normally distributed variables were shown as mean ± SD, while non-normally distributed ones were given as median (interquartile range). n (%) were used to summarize all categorical variables.
Variables showing a significant trend (P for trend < 0.001) were prioritized for subsequent analyses. Consequently, GDM, HDP, CD, and LGA were selected as primary outcomes. Spearman correlation analysis was conducted using the Correlation Plot function in Origin 2025 to explore relationships among the TyG index, related metrics, and metabolic variables. We also evaluated associations using univariate and multivariable logistic regressions, with odds ratios (OR) and 95% confidence intervals (CI) in three models. Model 1 used unadjusted raw data, Model 2 was adjusted for demographic and clinical factors, and Model 3 was further adjusted for metabolic markers. Covariates were as previously outlined. TyG and its derived indices were treated as categorical variables to evaluate independent risk factors for pregnancy outcomes in all models.
To evaluate the potential for nonlinear associations, a restricted cubic spline (RCS) model was employed with adjustment for all covariates. Receiver operating characteristic curves were constructed, and areas under the curve (AUCs) were computed to assess the predictive ability of the TyG index and its related indices for adverse maternal and neonatal outcomes. In addition, to confirm whether the TyG index and its derived indices significantly improved the predictive ability, the DeLong test was employed to compare the AUCs among the models. To assess potential effect modification, we stratified participants based on clinically established thresholds: Age was categorized by the advanced maternal age cut-off (< 35 vs ≥ 35 years)[27], and pre-pregnancy BMI was classified according to the criteria for overweight in Chinese adults (< 24 vs ≥ 24 kg/m²)[28]. Interaction effects were tested using the likelihood ratio test.
Data analysis was conducted using SPSS version 27.0 and R 4.2.1. Statistical significance was set at P < 0.05, with P < 0.01 indicating high significance.
RESULTS
Baseline characteristics
The study comprised 10422 pregnant women (Figure 1), with a mean age of 30.0 ± 3.3 years. Among them, 58.0% were primigravidae. Table 1 presents the baseline characteristics stratified by TyG index tertiles. With increasing TyG index tertiles, significant upward trends were observed in maternal age, pre-pregnancy BMI, SBP, DBP, WC, OGTT-0 hour glucose, OGTT-1 hour glucose, OGTT-2 hour glucose, TP, ALB, UA, TC, TG, LDL-C, HbA1c, TyG-BMI, TyG-WC, and TyG-WHtR (all P < 0.001). Conversely, downward trends were noted for ALT, AST, HDL-C and UREA (all P < 0.001). Similar patterns were observed when stratified by TyG-BMI, TyG-WC, and TyG-WHtR tertiles (Supplementary Tables 1-3).
Correlations between TyG index and its derived indices and glycemic/lipid parameters
Spearman correlation analysis was employed to explore the interrelationships between the TyG index, its derivatives, and a panel of glycemic and lipid markers (Figure 2). A consistent pattern of positive associations was observed for the TyG-based indicators with TC, TG, LDL-C, OGTT-0 hour, OGTT-1 hour, OGTT-2 hour, and HbA1c. Conversely, a significant inverse relationship was evident with HDL-C. All of these correlations were statistically significant (P < 0.001). Additionally, the four indices were also strongly intercorrelated. The Spearman coefficients between TyG and TyG-BMI, TyG-WC, and TyG-WHtR were 0.53, 0.47, and 0.47, respectively (all P < 0.001).
Figure 2 Spearman correlations between triglyceride-glucose index and its derived indices with glycemic and lipid parameters.aP < 0.05. bP < 0.01. cP < 0.001. TC: Total cholesterol; TG: Triglycerides; HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; OGTT: Oral glucose tolerance test; HbA1c: Glycated hemoglobin; TyG: Triglyceride-glucose index; BMI: Body mass index; WC: Waist circumference; WHtR: Waist-to-height ratio.
Relationship of TyG and its derived indices with maternal and neonatal outcomes
The incidence of adverse maternal and neonatal outcomes categorized by tertiles of TyG and its derived indices is shown in Supplementary Tables 4-7. The incidence of GDM, HDP, CD, and LGA was significantly higher in the T2 and T3 groups compared to the T1 group (P for trend < 0.001). No notable differences were observed between the groups for other outcomes, prompting further investigation into these four specific complications.
Both unadjusted and multivariable-adjusted logistic regression models (Table 2) showed significant associations between the TyG index and its derived indices and the four APOs and displayed a tendency to increase with the four indices. In the fully adjusted model (Model 3), when compared with women in the lowest tertile, those in the highest tertiles of TyG, TyG-BMI, TyG-WC, and TyG-WHtR were significantly associated with increased risks of both GDM and LGA. The adjusted ORs (95%CIs) for GDM were 1.90 (1.62-2.24), 1.20 (1.02-1.41), 1.50 (1.24-1.82) and 1.69 (1.40-2.04), respectively; and for LGA were 1.39 (1.17-1.64), 3.28 (2.75-3.91), 2.63 (2.14-3.23) and 1.83 (1.50-2.24), respectively. In Model 3, only TyG-BMI showed a significant association with HDP, with an adjusted OR of 1.46 (95%CI: 1.14-1.87) for the top tertile, and TyG showed no significant association with CD (P for trend = 0.215).
Table 2 Association of triglyceride-glucose index and its derived indices with adverse pregnancy outcomes.
RCSs analysis was conducted to further investigate the dose-response relationships between TyG-related indices, and four adverse outcomes, adjusting for multiple covariates (Figure 3). Regarding GDM, TyG, TyG-WC, and TyG-WHtR showed significant strong linear associations (all P for overall < 0.001; P for nonlinear > 0.05), while TyG-BMI was not significant (P for overall = 0.092). Consistent with the logistic regression analysis, only TyG-BMI was associated with HDP (P for overall < 0.001), with a linear trend (P for nonlinear = 0.949). Additionally, all indices were positively associated with CD and LGA (all P for overall < 0.05). The risks of CD demonstrated significant linear increases with rising levels of TyG, TyG-BMI, TyG-WC, and TyG-WHtR (Figure 3, all P for nonlinear > 0.05). For LGA, a linear increase was observed with TyG (P for nonlinear = 0.650) but non-linear associations were observed with TyG-BMI, TyG-WC, or TyG-WHtR (all P for nonlinear < 0.05) (Figure 3).
Figure 3 Dose-response relationship of triglyceride-glucose and its derived indices with the risk of gestational diabetes mellitus, hypertensive disorders of pregnancy, cesarean delivery, and large for gestational age.
A-D: Associations of triglyceride-glucose index (TyG), TyG-body mass index, TyG-waist circumference, and TyG-waist-to-height ratio with the risk of gestational diabetes mellitus; E-H: Associations of the four indices with the risk of hypertensive disorders of pregnancy; I-L: Associations of the four indices with the risk of cesarean delivery; M-P: Associations of the four indices with the risk of large for gestational age. GDM: Gestational diabetes mellitus; HDP: Hypertensive disorders of pregnancy; CD: Cesarean delivery; LGA: Large for gestational age; BMI: Body mass index; TyG: Triglyceride-glucose index; WC: Waist circumference; WHtR: Waist-to-height ratio.
Predictive capacity comparison
AUC values for TyG, TyG-BMI, TyG-WC, and TyG-WHtR were calculated, and all of them were significant predictors of GDM, HDP, CD, and LGA (Table 3 and Figure 4). For GDM, the AUC of TyG alone was 0.611 (0.596-0.626), comparable to TyG-WC [0.613 (0.596-0.629)] and TyG-WHtR [0.615 (0.598-0.631)]. TyG-BMI demonstrated the lowest AUC value of 0.594 (0.579-0.609). Compared with TyG, only TyG-WHtR showed statistically significant but minimal improvements in integrated discrimination improvement (IDI) (IDI = 0.003, P = 0.036) and continuous net reclassification improvement (cNRI) (cNRI = 0.064, P = 0.030). For HDP, TyG-BMI demonstrated the highest AUC of 0.667 (0.645-0.688), followed by TyG-WC [0.644 (0.620-0.668)] and TyG-WHtR [0.639 (0.616-0.663)], all of which outperformed the standalone TyG index [0.606 (0.584-0.628)]. Compared with TyG alone, TyG-BMI, TyG-WC, and TyG-WHtR showed significant improvements in predicting HDP, as evidenced by IDI (all P < 0.001) and cNRI (all P < 0.001). Furthermore, in predicting CD and LGA, all TyG-derived indices exhibited significant enhancement compared to TyG alone, with higher AUC values, IDI, and cNRI (all P < 0.001).
Figure 4 Receiver operating characteristic curves of triglyceride-glucose and its derived indices for predicting adverse maternal and neonatal outcomes.
A: Receiver operating characteristic (ROC) curves for gestational diabetes mellitus; B: ROC curves for hypertensive disorders of pregnancy; C: ROC curves for cesarean delivery; D: ROC curves for large for gestational age. GDM: Gestational diabetes mellitus; HDP: Hypertensive disorders of pregnancy; CD: Cesarean delivery; LGA: Large for gestational age; BMI: Body mass index; TyG: Triglyceride-glucose index; WC: Waist circumference; WHtR: Waist-to-height ratio; AUC: Areas under the curve.
Table 3 Comparative analysis of triglyceride-glucose index and its derived indices for predicting adverse pregnancy outcomes.
Subgroup analyses based on age and pre-pregnancy BMI were conducted (Supplementary Tables 8-11). After adjusting for confounders in Model 3, consistent associations were observed across subgroups between the tertiles of the TyG index and its derivatives and the risks of GDM and LGA (P for interaction < 0.05). However, a significant interaction was found for age, with a stronger association between the indices and HDP in women aged ≥ 35 years, whereas the association with CD risk was more pronounced in women aged < 35 years (both P for interaction < 0.05).
DISCUSSION
Overall, the TyG index was moderately correlated with its derived indices. Elevated levels of TyG, TyG-BMI, TyG-WC, and TyG-WHtR were independently associated with increased risks of GDM, HDP, CD, and LGA after adjustment. Dose-response analysis confirmed predominantly linear associations for most outcomes, though notable non-linear relationships emerged for LGA. Particularly, for GDM, TyG-WHtR showed statistically significant but minimal improvements compared to TyG alone. Furthermore, all TyG-derived indices significantly outperformed TyG alone in predicting HDP, CD, and LGA, with higher AUC, IDI, and cNRI values. The robustness of findings was further validated through subgroup analyses. Collectively, these findings indicate that the TyG index and its obesity parameter-integrated derivatives provide enhanced predictive markers for a spectrum of APOs.
Mechanistically, exacerbated IR elevates maternal blood glucose levels[29], thereby contributing to the development of GDM. Maternal hyperglycemia further induces fetal hyperinsulinemia, accelerating fetal adipogenesis and protein synthesis, which in turn promotes macrosomia and LGA[1,30]. Concurrently, aggravated IR during pregnancy may lead to endothelial dysfunction, placental insufficiency, and metabolic disorders (e.g., reduced nitric oxide, elevated oxidative stress, dyslipidaemia, and prostaglandin E2 inhibition), collectively causing pregnancy-induced hypertension[31-34]. Epidemiologically, IR significantly increases the risks of GDM, HDP and macrosomia/LGA[35-38]. These complications synergistically elevate the likelihood of CD. Genetically, IR has been demonstrated to have a causal relationship with GDM[39], pregnancy-induced hypertension[40], and macrosomia[41].
The TyG index is an economical, readily accessible, and reliable assessment tool[12]. Besides, body fat content and its distribution are closely associated with IR. Studies have confirmed that adipose tissue functions as an endocrine organ, secreting hormones and cytokines that bidirectionally regulate glucose homeostasis and insulin signaling[42]. Given the interplay between IR and obesity, an increasing number of studies have investigated whether combining the TyG index with obesity-related parameters (such as BMI, WC, and WHtR) could enhance risk stratification for APOs[43,44]. This study revealed moderate correlations linking the TyG metric to its variations (TyG-BMI, TyG-WC, and TyG-WHtR) (correlation coefficients: 0.47-0.53; Figure 2), while the interconnections among the derived TyG indices reached 0.62-0.96. These results indicate the potential of the TyG index and its derivatives in the evaluation of distinct dimensional characteristics of IR.
Given the heterogeneity in study populations, outcome definitions, and gestational weeks at assessment, the predictive utility of the TyG index and its derivatives for APOs remains controversial. Large-scale research from the National Health and Nutrition Examination Survey[16], China[45], and South Korea[17] has demonstrated that higher TyG levels are associated with an increased risk of GDM, which aligns with our findings. However, Sánchez-García et al[19] found that the TyG index did not show significant GDM predictive ability in Latin American pregnant women. Meng et al[44] reported that TyG-BMI was superior to the TyG index alone in predicting GDM, HDP, and LGA. Specifically, while Pan et al[33] found that elevated TyG levels were significantly associated with the prevalence of gestational hypertension, a prospective study by Li et al[46] including 11387 pregnant women in southeastern China revealed that although an association between first-trimester TyG and gestational hypertension risk was observed in univariate analysis, this correlation lost statistical significance following adjustment for potential confounders. Furthermore, a recent prospective cohort study conducted in Mexico City failed to identify any independent association between the TyG index and LGA, SGA, preterm birth, or CD[47].
Given these inconsistent findings and the limited research on derived indices, our study systematically evaluated and compared the associations of TyG, TyG-BMI, TyG-WC, and TyG-WHtR with GDM, HDP, CD, and LGA based on a large-scale cohort in Eastern China. Following adjustment for potential covariates, all four indices demonstrated significant positive correlations with the aforementioned adverse outcomes, with ORs increasing across tertiles (Table 2). Regarding GDM prediction, our results indicated that the area under the curve (AUC) for TyG was 0.611 (0.596-0.626), slightly lower than that of TyG-WHtR [0.615 (0.598-0.631)], which is consistent with previous studies[16,47]. Our study confirmed that TyG-derived indices (TyG-BMI, TyG-WC, and TyG-WHtR) exhibited superior predictive performance for HDP, CD, and LGA compared to TyG alone. These findings align with early-pregnancy data reported by Meng et al[44]. Our RCS analysis revealed significant non-linear dose-response relationships between maternal TyG-BMI, TyG-WC, TyG-WHtR levels and LGA risk (Figure 3), corroborating the observations by Pan et al[33]. Subgroup analyses revealed stronger associations with HDP in women ≥ 35 years, likely reflecting the established role of advanced maternal age as a risk factor[48]. However, stronger associations with CD were observed in women < 35 years, possibly due to the larger sample size and greater statistical power in this subgroup (Supplementary Tables 8-11).
The underlying mechanisms for this differential predictive performance remain incompletely understood but may involve synergistic effects of IR and abdominal obesity in promoting LGA. TyG reflects systemic insulin sensitivity impairment, whereas WC and WHtR indicate visceral fat accumulation, both of which have demonstrated significant associations with chronic inflammation, endothelial dysfunction, and increased oxidative stress. These processes collectively contribute to APOs. Therefore, composite parameters integrating TyG with central obesity indices may offer superior predictive value over TyG alone.
The TyG index and its derivatives are promising candidates due to their low cost and ease of calculation. The primary contribution of our work is that it is among the first studies to systematically report and compare the independent associations between mid-pregnancy TyG index and its multiple obesity-derived indices with a spectrum of APOs. By leveraging a large, well-characterized cohort and standardized research methods, our study not only confirms these associations but, importantly, also elucidates a potential synergistic effect between IR and obesity, adding a new layer of understanding to the underlying pathophysiology.
This study has several limitations. First, due to its retrospective design, detailed information on maternal diets and lifestyle habits was lacking. Furthermore, since clinical data were retrieved from historical records, the possibility of inter-observer variation in data collection cannot be entirely ruled out. However, extensive covariate adjustments likely minimized residual confounding. Second, the influence of therapeutic interventions was not fully accounted for. Specifically, for participants diagnosed with GDM or HDP, standard management strategies (such as medical nutrition therapy or medication) were implemented according to clinical guidelines. These interventions could improve pregnancy outcomes, potentially attenuating the observed associations. Third, as the majority of pregnant women underwent examinations at community health centers rather than hospitals during the first trimester, analyses were confined to mid-pregnancy TyG and its derivatives, biochemical indicators, and anthropometric measurements. The lack of continuous dynamic monitoring during preconception and early pregnancy periods may compromise the comprehensiveness and temporal continuity of our observations. Finally, despite the considerable sample size, the study population was recruited from a single center in Eastern China. Therefore, extrapolating these findings to wider populations requires caution, and additional validation in varied geographic and healthcare contexts is necessary. Despite its limitations, this study is the first to systematically link mid-pregnancy TyG indices to adverse outcomes, offering crucial insights for future research and novel clinical risk assessment strategies. Future studies should explore the impact of pre-pregnancy and early-pregnancy TyG and its derivatives on pregnancy outcomes, as well as investigate potential causal mechanisms and feasible public health interventions.
CONCLUSION
In conclusion, the TyG index and its derivatives incorporating obesity parameters (TyG-BMI, TyG-WC and TyG-WHtR) were significantly associated with GDM, HDP, CD, and LGA. The TyG-derived indices exhibited improved predictive capability over the conventional TyG index in risk stratification and prediction of HDP, CD, and LGA.
Wang H, Li N, Chivese T, Werfalli M, Sun H, Yuen L, Hoegfeldt CA, Elise Powe C, Immanuel J, Karuranga S, Divakar H, Levitt N, Li C, Simmons D, Yang X; IDF Diabetes Atlas Committee Hyperglycaemia in Pregnancy Special Interest Group. IDF Diabetes Atlas: Estimation of Global and Regional Gestational Diabetes Mellitus Prevalence for 2021 by International Association of Diabetes in Pregnancy Study Group's Criteria.Diabetes Res Clin Pract. 2022;183:109050.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 964][Cited by in RCA: 777][Article Influence: 194.3][Reference Citation Analysis (0)]
Lowe WL Jr, Scholtens DM, Kuang A, Linder B, Lawrence JM, Lebenthal Y, McCance D, Hamilton J, Nodzenski M, Talbot O, Brickman WJ, Clayton P, Ma RC, Tam WH, Dyer AR, Catalano PM, Lowe LP, Metzger BE; HAPO Follow-up Study Cooperative Research Group. Hyperglycemia and Adverse Pregnancy Outcome Follow-up Study (HAPO FUS): Maternal Gestational Diabetes Mellitus and Childhood Glucose Metabolism.Diabetes Care. 2019;42:372-380.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 465][Cited by in RCA: 414][Article Influence: 59.1][Reference Citation Analysis (0)]
Zheng H, Xu M, Yang J, Xu M. Association between the triglyceride-glucose index and its combined obesity indicators and the risk of hypertension in middle-aged and older Chinese adults: A nationwide cross-sectional study.PLoS One. 2025;20:e0316581.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 3][Reference Citation Analysis (0)]
International Association of Diabetes and Pregnancy Study Groups Consensus Panel; Metzger BE, Gabbe SG, Persson B, Buchanan TA, Catalano PA, Damm P, Dyer AR, Leiva Ad, Hod M, Kitzmiler JL, Lowe LP, McIntyre HD, Oats JJ, Omori Y, Schmidt MI. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy.Diabetes Care. 2010;33:676-682.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 2777][Cited by in RCA: 3388][Article Influence: 211.8][Reference Citation Analysis (3)]
Zhu YT, Xiang LL, Chen YJ, Zhong TY, Wang JJ, Zeng Y. Developing and validating a predictive model of delivering large-for-gestational-age infants among women with gestational diabetes mellitus.World J Diabetes. 2024;15:1242-1253.
[PubMed] [DOI] [Full Text]
Hauth JC, Clifton RG, Roberts JM, Myatt L, Spong CY, Leveno KJ, Varner MW, Wapner RJ, Thorp JM Jr, Mercer BM, Peaceman AM, Ramin SM, Carpenter MW, Samuels P, Sciscione A, Tolosa JE, Saade G, Sorokin Y, Anderson GD; Eunice Kennedy Shriver National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network. Maternal insulin resistance and preeclampsia.Am J Obstet Gynecol. 2011;204:327.e1-327.e6.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 79][Cited by in RCA: 84][Article Influence: 5.6][Reference Citation Analysis (0)]
Alers RJ, Ghossein-Doha C, Brandt Y, Kooi ME, Gerretsen SC, Jansen JFA, Backes WH, van de Ven V, Hurks PPM, Spaanderman MEA. Associations of Metabolic Syndrome and Insulin Resistance With Attenuated Executive Function Post-Preeclampsia: A Nested Case-Control Study.BJOG. 2025;132:1269-1277.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in RCA: 1][Reference Citation Analysis (0)]
Gurza G, Martínez-Cruz N, Lizano-Jubert I, Arce-Sánchez L, Suárez-Rico BV, Estrada-Gutierrez G, Montoya-Estrada A, Romo-Yañez J, Solis-Paredes JM, Torres-Torres J, González-Ludlow I, Rodríguez-Cano AM, Tolentino-Dolores M, Perichart-Perera O, Reyes-Muñoz E. Association of the Triglyceride-Glucose Index During the First Trimester of Pregnancy with Adverse Perinatal Outcomes.Diagnostics (Basel). 2025;15:1129.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in RCA: 9][Reference Citation Analysis (0)]
