Association of the triglyceride glucose-body roundness index with stroke risk in middle-aged and elderly Chinese adults without diabetes: a prospective cohort study

Introduction

Stroke remains a leading global cause of mortality, responsible for 7.3 million deaths in 2021 and accounting for 10.7% of total worldwide mortality (1,2). In China, stroke has emerged as the foremost cause of death and disability, representing a critical public health challenge as it contributes to 44% of all adult deaths (3).

Insulin resistance (IR), a key driver of cardiovascular disease (CVD), has gained attention as a potential indicator of stroke risk (4). The triglyceride-glucose index (TyG), an established surrogate marker of IR, along with its derivatives that combine the TyG index with waist circumference (WC), body mass index (BMI), and waist-to-height ratio (WHtR), have demonstrated strong associations with arterial stiffness and stroke incidence (5). Evidence from population-based studies indicated that elevated baseline and long-term cumulative mean TyG index levels independently confer an increased hazard for incident stroke and ischemic stroke (IS) (6). A higher TyG index is significantly associated with elevated risks of stroke and fatality (7). Additionally, the data suggested a clear and complex interaction between the TyG index and visceral obesity in the association with stroke risk (8). This finding aligns with evidence that the content and distribution of body fat (especially visceral adipose tissue) are closely related to IR, CVD, and cardiovascular mortality (9). Visceral obesity is primarily characterized by the excessive accumulation of intra-abdominal fat depots. The body roundness index (BRI) serves as a robust surrogate measure of visceral adiposity, derived from the height-to-waist circumference (WC) ratio (10). This methodological approach allows the BRI to more accurately reflect visceral fat accumulation and distribution than traditional measures like BMI and raw WC (11). Previous studies have consistently established a significant link between BRI and an increased risk of hypertension, diabetes, dyslipidemia, and CVD (9,10,12). In light of the documented mechanistic link between IR and excess adiposity, recent investigations have increasingly prioritized assessing whether the interaction of the TyG index and anthropometric obesity parameters can optimally improve risk assessment for incident stroke (13,14). Some studies proposed that this combination provides a more accurate risk stratification than the TyG index alone (15), while inconsistent results have been documented by others (16,17). While the dichotomization of TyG and BRI based on median values provided initial insights, such artificial grouping constrained the ability to capture the nature of metabolic-obesity phenotypes (18). The TyG-BRI index has emerged as a superior, integrated metric that offers a more operable and quantitative approach to risk stratification by harmonizing metabolic and morphological dimensions. Nevertheless, empirical evidence regarding their joint impact on incident stroke remains limited. Furthermore, as a major independent risk factor for stroke onset, diabetes presents a significant confounding variable when exploring the interactive or synergistic effects of TyG and BRI on stroke risk in the general population (19,20).

The primary objective of this prospective cohort study is to evaluate the associations of the TyG index, BRI, and their composite measure (TyG-BRI) with the risk of incident stroke among middle-aged and elderly non-diabetic Chinese adults. Furthermore, we aim to elucidate the potential interaction between TyG and BRI to determine whether their joint evaluation significantly improves prognostic risk stratification for stroke events. We present this article in accordance with the STROBE reporting checklist (available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-477/rc).

Methods

Study design and study population

Utilizing the China Health and Retirement Longitudinal Study (CHARLS) (21), we performed a prospective analysis within a nationwide sample comprising Chinese individuals aged 45 and above. The cohort employed a multistage stratified probability sampling approach. The baseline survey recruited 17,705 participants from 150 counties or districts and 450 urban communities or villages across 28 provinces in China, ensuring a robust and representative sample. The baseline survey was conducted in May 2011, with follow-up assessments performed every two to three years thereafter. For the current study, eligibility was restricted to participants aged 45 years and older. After excluding participants with baseline histories of stroke, cancer or diabetes, as well as those with missing follow-up data on stroke incidence, 14,079 individuals remained eligible for subsequent analysis. A further 6,683 participants were excluded due to missing triglyceride (TG) and fasting blood glucose (FBG) data. Additional exclusions included incomplete baseline characteristics and a follow-up duration of less than two years. Ultimately, 7,396 participants aged ≥45 years without diabetes were enrolled from the baseline survey conducted between May 2011 and March 2012. The systematic recruitment and exclusion process for the study population was detailed in Figure 1. The CHARLS study was granted ethical approval by the Institutional Review Board of Peking University (approval No. IRB00001052-11015 for household survey and No. IRB00001052-11014 for blood sample), and all participants provided informed written consent. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Figure 1 The flowchart of study participants. BRI, body roundness index; CHARLS, The China Longitudinal Study of Health and Retirement; FBG, fasting blood glucose; TG, triglycerides; TyG, triglyceride glucose index; WC, waist circumference.

Assessment of exposures

Venous specimens were obtained after a minimum 8-hour fasting period and subsequently transferred to Beijing for specialized analysis. The Clinical Laboratory of Capital Medical University, which holds accreditation from the Beijing Health Bureau, quantified TG and FBG levels via enzymatic colorimetric techniques. Both parameters demonstrated excellent reproducibility, evidenced by coefficients of variation that remained within a strict 5% threshold. The calculation of IR indices and obesity-related parameters was calculated using the formula:

BRI=364.2−365.5×1−(WC/2π0.5×height)2[1]

TyG=lnfastingtriglyceride(mg/dL)×fastingglucose(mg/dL)2[2]

TyG-BMI=TyG×BMI[3]

TyG-WC=TyG×WC(cm)[4]

TyG-WHtR=TyG×WC(cm)height(cm)[5]

TyG-BRI=TyG×BRI[6]

Assessment of incident stroke

We established incident stroke as the central outcome of interest, covering the period up to the fifth survey wave in 2020. Data acquisition was conducted by medically proficient investigators who employed a standardized inquiry to verify any history of physician-diagnosed stroke among the respondents: “Have you been diagnosed with a stroke by a physician?”. Subsequently, the research team conducted data verification and validation procedures to ensure accuracy (22). The investigators ascertained the time of first stroke diagnosis (including hemorrhagic and IS) by a medical institution, based on inquiries to the participants or their family members.

Definitions of covariates

The venous blood samples were collected by trained professionals using a standard phlebotomy process and tested for biochemical parameters. Following the separation of the venous blood into plasma and buffy coat, the plasma was kept in three 0.5 mL cryovials, while the buffy coat was kept in a different cryovial. After attaining a stable temperature of −20 ℃ through immediate freezing, the cryovials were transported to the centralized laboratory at the China CDC in Beijing. This inter-facility shipment was executed within a fortnightly period to ensure biochemical stability (22). The blood samples were tested for complete blood count (CBC) and blood chemistry panel. The methods and detection limits of bioassays were shown in Appendix 1. Laboratory tests included measurements of total cholesterol (TC), TG levels, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), glycosylated hemoglobin A1c (HbA1c), uric acid (UA), and C-reactive protein (CRP). Additionally, the information about sociodemographic characteristics (age, sex, place of residence, marital status, educational background), anthropometric measurements [BMI, systolic blood pressure (SBP), and diastolic blood pressure (DBP)], lifestyle factors (smoking and drinking), and Health-related variables (history of diabetes, hypertension, heart disease, kidney disease, liver disease, chronic lung diseases and dyslipidemia) was collected by trained interviewers with a structured questionnaire. Diabetes was defined as fasting plasma glucose ≥7.0 mmol/L or/and HbA1c ≥6.5%, no self-reported diagnosis, or antihyperglycemic medication use. Hypertension was defined as SBP ≥140 mmHg, DBP ≥90 mmHg, current use of antihypertensive medication, or self-reported history of hypertension. Dyslipidemia was defined as TC levels ≥240 mg/dL, TG ≥200 mg/dL, LDL-C levels ≥160 mg/dL, HDL-C levels <40 mg/dL, current use of lipid-lowering medication, or self-reported history of dyslipidemia. The detailed information about the categorized characteristics was presented in Table S1.

Statistical analysis

The baseline characteristics were summarized according to the quartile groups of BRI, TyG, and TyG-BRI. Continuous variables adhering to normal distributions were expressed as means with standard deviations (SDs), whereas those exhibiting skewed distributions were presented as medians with interquartile ranges (IQRs). Categorical data were summarized using frequencies and percentages. The differences among the groups were analyzed through the analysis of variance (ANOVA) or Kruskal-Wallis rank sum test for continuous variables and chi-square test or Fisher’s exact test for categorical variables. To determine the distribution of continuous data, we have employed both Q-Q plots for visual assessment and D’Agostino’s K² test, specifically targeting symmetry (skewness) and peakedness (kurtosis) to verify the assumption of normality.

Cox proportional hazards regression models were employed to estimate hazards ratio (HRs) and 95% confidence intervals (CIs) for the primary outcome of new-onset stroke. Associations were assessed for BRI, TyG, and the combined TyG-BRI index, with each variable analyzed both in continuous form and by quartiles. Crude was an unadjusted model; model 1 was adjusted for sex and age; model 2 was adjusted for model 1 plus BMI, education, marital status, residence, living standard, smoking status, drinking status, UA, serum creatinine, CRP, FBG, SBP, DBP; Model 3 was adjusted as model 2 plus history of hypertension, dyslipidemia, liver disease, kidney disease, chronic lung diseases. To investigate the dose-response relationships of BRI, TyG, and TyG-BRI with stroke incidence, a restricted cubic spline (RCS) based on Cox regression models was employed, adjusting covariates in model 3. The population was stratified according to the TyG reference values of the RCS curves, and Cox regression with the same model as above was used to explore the relationship between BRI and stroke incidence in different stratifications separately.

In the sensitivity analysis, individuals with diabetes were included to evaluate the robustness of the primary findings across a broader study population. The relationships between the continuous values and quartiles of TyG-BRI, BRI, and TyG and the risk of stroke were re-examined. Additionally, stratified analysis was performed based on different levels of TyG.

Several subgroup analyses were performed to assess the potential heterogeneity in the association between the TyG-BRI, BRI and TyG index on stroke incidence. The categorized confounders include age (<60 and ≥60 years), gender, smoking, drinking, BMI (<24 and ≥24 kg/m²), hyperlipidemia, hypertension, liver disease, kidney disease, chronic lung disease. Meanwhile, we explored their interacting relationships.

To evaluate additive interactions, we employed three complementary metrics: the relative excess risk due to interaction (RERI), the attributable proportion (AP) of interaction, and the synergy index (SI) (23). Each metric quantifies distinct dimensions of interaction—RERI reflects the excess risk attributable to interaction, AP represents the proportion of joint effects explained by interaction, and SI measures the ratio of combined effects to independent effects. The absence of interaction between BRI and TyG in relation to stroke incidence is indicated by RERI =0, AP =0, and SI =1. A synergistic interaction, where the combined effect exceeds the sum of individual effects, is demonstrated when RERI >0, AP >0, and SI >1. Conversely, an antagonistic interaction (combined effect weaker than the sum of individual effects) is inferred when RERI <0, AP <0, and SI <1. 95% CIs for these metrics were derived using the delta method to assess statistical significance.

We assessed the performance of BRI, TyG, TyG-BRI index, and TyG-BMI by evaluating their discriminative capabilities using the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC) (24). DeLong’s method for comparing the differences in the AUC (25). Additionally, we explored the optimizing effect of the TyG-BRI index on the predictive model by comparing the predictive abilities of the upgraded model and the basic model for stroke risk, and analyzed the differences in the AUC using the DeLong method. The basic model adjusted for sex and age, BMI, education, marital status, residence, living standard, smoking status, drinking status, UA, serum creatinine, CRP, FBG, SBP, DBP, TC, HDL, TG, LDL, obesity. All statistical analyses were conducted using R statistical software, version 4.4.0 (R Foundation). A two-sided P value of less than 0.05 was considered to indicate statistical significance.

Results

Baseline characteristics of participants

The 14,079 participants remained eligible for further study, 6,683 participants were excluded without data on TG, and FBG. People were also excluded if their baseline characteristic was incomplete or if they had followed up less than two years. From the baseline survey (May 2011 to March 2012), we included 7,396 participants aged ≥45 years without diabetes.

A total of 6,315 non-diabetic participants (47.0% male) were enrolled in the TyG-BRI study, with a mean age (SD) of 58.2±9.7 years. Participants were categorized into four subgroups based on TyG-BRI quartiles and the baseline characteristics of the participants were summarized in Table 1. Statistically significant differences were observed in most baseline characteristics among the four subgroups, including age, sex, living standard, smoking status, drinking, and physical examinations, CRP, FBG, TC, TG, LDL-C, HDL-C, HbA1c, estimated glomerular filtration rate (eGFR), hypertension, hyperlipidemia, kidney disease, chronic lung disease. Additionally, the baseline characteristics of participants accord­ing to the quartiles of BRI and TyG were presented in Tables S2,S3.

Association between the TyG, BRI, and TyG-BRI and risk of stroke

To maximize the inclusion of eligible participants, individuals were excluded on an index-specific basis only if they lacked the necessary parameters for calculation. Consequently, this approach resulted in varying sample sizes for exploring the associations between TyG (n=7,396), BRI (n=6,862), and TyG-BRI (n=6,315) with stroke incidence, respectively. During a median follow-up period of 6.5 years, there were 440 cases of stroke were recorded for TyG-BRI analysis. In fully adjusted models, elevated TyG-BRI (Q4 vs. Q1: HR =2.29, 95% CI: 1.58–3.33), BRI (Q4 vs. Q1: HR =2.12, 95% CI: 1.49–3.02), and TyG (Q4 vs. Q1: HR =1.82, 95% CI: 1.30–2.53) were significantly associated with higher stroke risk (P trend <0.001, P trend <0.001, P-trend =0.01). Meanwhile, the continuous TyG-BRI, BRI, and TyG were strongly related to stroke incidence (per SD increased P<0.001, P<0.001, P=0.004) (Table 2). Stratified analysis by TyG levels revealed a significant positive association between BRI and stroke incidence in the high-TyG (≥8.37) population (P<0.001), whereas no statistically significant correlation was observed in the low-TyG group (<8.37) (Table 3). These RCS curves suggested a significant and linear relationship between TyG-BRI and stroke incidence (all P for overall <0.001 and P for non-linear =0.084) (Figure 2).

Figure 2 Linear dose-response associations of TyG-BRI, BRI, and TyG with stroke incidence. (A-C) Graphs show HRs for stroke incidence based on Cox hazards regression model 3 (adjusted for sex and age, BMI, education, marital status, residence, living standard, smoking status, drinking status, uric acid, serum creatinine, CRP, FBG, SBP, DBP, history of hypertension, dyslipidemia, liver disease, kidney disease, chronic lung diseases). Solid lines indicate HRs. Shadow shapes indicate 95% CIs. BMI, body mass index; BRI, body roundness index; CI, confidence interval; CRP, C-reactive protein; DBP, diastolic blood pressure; FBG, fasting blood glucose; HR, hazard ratio; SBP, systolic blood pressure; TyG, triglyceride glucose index.

Sensitivity analysis

To enhance the robustness of study findings, we performed a sensitivity analysis by including the participants with diabetes. The results demonstrated that TyG-BRI (Q4 vs. Q1: HR =1.51, 95% CI: 1.13–2.02), BRI (Q4 vs. Q1: HR =1.52, 95% CI: 1.16–2.00), and TyG (Q4 vs. Q1: HR =1.40, 95% CI: 1.11–1.76) remained significantly and positively associated with stroke risk (all P trend <0.001, Table S4). Furthermore, when stratified by TyG levels, BRI showed a significant positive correlation with stroke risk in both subgroups (TyG ≥8.37, Q4 vs. Q1: HR =2.08, 95% CI: 1.11–3.91; TyG <8.37, Q4 vs. Q1: HR =1.94, 95% CI: 1.16–3.24, Table S5).

Subgroup analysis

Subgroup-specific associations were scrutinized to determine if the impact of TyG-BRI on stroke incidence varied by baseline characteristics. These exploratory analyses confirmed that the hazard patterns were consistent across most demographic and clinical cohorts, aligning with our overarching conclusions (Figure 3). No multiplicative interactions were detected between TyG-BRI quartiles and any stratification factors (Figure S1). Additionally, the study did not reveal the interaction of multiple factors on the relationship between BRI, TyG and stroke risk (Figures S2-S5).

Figure 3 Subgroup and interaction analyses among the TyG-BRI and stroke incidence across various subgroups. BRI, body roundness index; CI, confidence interval; HR, hazard ratio; TyG, triglyceride glucose index.

The interactive effect of the TyG index and BRI on stroke incidence

After adjusting comprehensively for potential confounders, we found that the 95% CIs for the RERI and AP included 0, while those for the SI and the multiplicative effect included 1. These findings suggest that there were no statistically significant additive or multiplicative interactions between the TyG index and BRI in relation to stroke (Table 4).

Incremental predictive performance of TyG-BRI in the incident stroke

Analysis of the ROC curves revealed that the integrated TyG-BRI index outperformed other metrics in stroke risk prediction, yielding the most robust AUC (AUC: 0.60; 95% CI: 0.57–0.62), followed by BRI and TyG-BMI (Figure 4A). Our baseline predictive frameworks were developed by extending a core model, which accounted for age and sex, to encompass a comprehensive suite of sociodemographic factors, lifestyle behaviors, and clinical biomarkers, including renal function, lipid profiles, and systemic inflammatory markers. The TyG-BRI index improved the predictive capacity of the base model for stroke incidence (C-statistics: 0.677 vs. 0.640, P=0.02) (Figure 4B). Despite the inclusion of hypertension in the basic model, which did enhance its predictive capability for stroke incidence, it remains inferior in strength compared to our capacity to incorporate TyG-BRI.

Figure 4 The receiver operating characteristic curves of BRI and TyG to predict stroke incidence. (A) Comparison of the predictive value of TyG-BRI with other indices for stroke incidence; (B) comparison of the predictive value of TyG-BRI model with basic model for stroke incidence. The basic model adjusted for sex and age, BMI, education, marital status, residence, living standard, smoking status, drinking status, uric acid, serum creatinine, CRP, FBG, SBP, DBP, TC, HDL-C, TG, LDL-C, obesity. AUC, area under the curve; BMI, body mass index; BRI, body roundness index; CI, confidence interval; CRP, C-reactive protein; DBP, diastolic blood pressure; FBG, fasting blood glucose; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; TyG, triglyceride glucose index.

Discussion

In this prospective cohort study, we established a significant and robust association between the novel TyG-BRI composite index and the risk of stroke. Importantly, the TyG-BRI exhibited superior prognostic accuracy for stroke risk compared to the TyG index or the BRI utilized individually. While formal statistical testing did not detect a significant interaction between TyG and BRI, the TyG-BRI composite consistently enhanced the overall predictive capability of the risk model. Collectively, these findings underscore the potential of TyG-BRI to serve as a clinically valuable composite biomarker for stroke risk stratification in this non-diabetic population.

The TyG index was calculated as a product of fasting TG and glucose, and is a widely validated and readily accessible surrogate measure for IR (26). Mechanistically, the elevation of TyG was linked to a cascade of atherogenic processes, including chronic low-grade inflammation, endothelial dysfunction, and accelerated dyslipidemia (27,28). Consequently, extensive epidemiological evidence consistently demonstrates that elevated TyG levels were significantly associated with risks of CVDs and stroke (16). However, the clinical manifestation of IR is profoundly heterogeneous, its impact being substantially modulated by the individual’s adiposity profile. Previous research has explored various TyG-anthropometric combinations, such as TyG-BMI, TyG-WC, TyG-WHtR, and TyG-ABSI (29). While these indices generally improve risk prediction over their components, each has constraints (30). TyG-BMI inherits the limitations of BMI regarding body composition. TyG-WC and TyG-WHtR effectively integrate IR with central obesity but may not fully capture the three-dimensional geometry of visceral fat (31). TyG-ABSI, though focused on body shape, may be less sensitive to the volumetric aspects of adiposity (14). Our results indicate that the TyG-BRI index outperforms these alternatives in its association with stroke risk. The BRI’s foundation in elliptical body modeling appears to offer a more accurate approximation of visceral fat volume, making its combination with TyG a particularly potent indicator of the high-risk “metabolically obese” phenotype (32). A particularly salient observation is the robust association of the TyG-BRI index with stroke risk among non-diabetic individuals. This population represents a crucial target for primary prevention, as they often fall into an intermediate-risk category where traditional diabetic thresholds are not met, yet subclinical metabolic and adiposity-related pathologies are actively progressing (33). In this context, TyG-BRI serves as a sensitive tool to identify those at elevated risk who might otherwise be missed by conventional glycemic criteria (30).

This prospective study provides novel evidence on the associations of TyG, BRI, and their composite marker (TyG-BRI) with incident stroke in non-diabetic Chinese adults. Our findings demonstrate that both elevated TyG and higher BRI independently predict stroke risk, consistent with prior studies linking metabolic dysregulation and visceral obesity to stroke incidence (34). Notably, the TyG-BRI index exhibited superior predictive performance compared to either marker alone, suggesting that the combined assessment of IR and body fat distribution may enhance early stroke risk stratification. The previous study lacked exploration of the clinical cut-off values for BRI and the TyG index (35,36). The use of median values to classify participants may not accurately reflect clinically meaningful threshold distinctions. Future research needs to establish more precise grouping criteria. Moreover, the study focused on the general population rather than a healthy population without severe metabolic diseases. Additionally, the study did not further explore the predictive role of the TyG-BRI combination for stroke risk (18). The enhanced predictive capacity of TyG-BRI may stem from its dual representation of pathogenic mechanisms. IR promotes endothelial dysfunction and prothrombotic states (37), while visceral adiposity drives systemic inflammation and atherogenic lipid profiles-processes that likely interact to accelerate cerebrovascular damage (38). Our results align with emerging evidence that integrative indices outperform single biomarkers in cardiovascular prognostication (14). The TyG-BRI index may represent a novel biomarker combining IR and obesity indices for stroke risk, following other composite indicators such as TyG-ABSI, TyG-CVAI, TyG-BMI, TyG-WC, and TyG-WHtR (14,39,40). The inverse relationship between the highest TyG-BRI quartile and chronic kidney disease (CKD) prevalence may be partly attributable to reverse causality. In this context, metabolic alterations and weight loss associated with undiagnosed CKD could reduce measured adiposity, consistent with the obesity paradox. Furthermore, unmeasured confounding factors, such as earlier use of renoprotective medications and more intensive clinical monitoring in high-risk patients, might attenuate the expected positive association (41).

Advantages of study

This study systematically evaluates the predictive value of the TyG-BRI composite index for stroke incidence in a non-diabetic Chinese population, providing new evidence for early identification of metabolic risk factors. The prospective cohort design, along with long-term follow-up for endpoint events, strengthens the reliability of the findings. Furthermore, we explored the combined association between IR (TyG) and visceral obesity (BRI), offering a novel theoretical perspective on stroke pathogenesis.

Study limitations

Several limitations should be acknowledged in this study. First, despite multivariate adjustments, residual confounding from unmeasured variables such as dietary patterns may still exist. Second, the observational nature of the design limits causal inference, highlighting the need for mechanistic studies to clarify how the interplay between IR and adiposity contributes to stroke pathogenesis. Additionally, in the CHARLS cohort, the timing of outcome events was based on self-reported questionnaires, which may introduce recall bias and preclude precise determination of the exact onset time. It should also be noted that our study population was limited to middle-aged and elderly Chinese adults; therefore, caution is warranted when generalizing these findings to other ethnic groups or younger individuals.

Conclusions

Our investigation established that both the TyG index and BRI were significantly associated with the risk of stroke within the non-diabetic population. While formal statistical testing did not yield a significant interactive effect between the two indices, their composite measure (TyG-BRI) exhibited superior prognostic accuracy for stroke risk compared to either TyG or BRI utilized in isolation. The TyG-BRI index significantly enhanced the discriminatory power of the baseline clinical risk model. Collectively, these results underscore the potential clinical utility of the TyG-BRI index for optimizing the identification of high-risk individuals and for developing targeted monitoring for stroke prevention.

Acknowledgments

We sincerely thank all members of the CHARLS team for their efforts in data collection and all participants for their contributions. We also gratefully acknowledge Professor Evropi Theodoratou from University of Edinburgh, for her guidance on this study and her assistance in language refinement.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-477/rc

Peer Review File: Available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-477/prf

Funding: This work was supported by Zhejiang Provincial Health Industry Science and Technology Program (No. 2025HY0058 to J.J.); and the Natural Science Foundation of Zhejiang Province (No. LZ24H310002 to H.B.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://cdt.amegroups.com/article/view/10.21037/cdt-2025-477/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The CHARLS study was granted ethical approval by the Institutional Review Board of Peking University (approval No. IRB00001052-11015 for household survey and No. IRB00001052-11014 for blood sample), and all participants provided informed written consent. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Global, regional, and national burden of household air pollution, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 2025;405:1167-81. [Crossref] [PubMed]
2. Feigin VL, Brainin M, Norrving B, et al. World Stroke Organization: Global Stroke Fact Sheet 2025. Int J Stroke 2025;20:132-44. [Crossref] [PubMed]
3. Tu WJ, Wang LDSpecial Writing Group of China Stroke Surveillance Report. China stroke surveillance report 2021. Mil Med Res 2023;10:33. [Crossref] [PubMed]
4. Huang H, Xiong Y, Zhou J, et al. The predictive value of estimated glucose disposal rate and its association with myocardial infarction, heart failure, atrial fibrillation and ischemic stroke. Diabetes Obes Metab 2025;27:1359-68. [Crossref] [PubMed]
5. Yang Y, Huang X, Wang Y, et al. The impact of triglyceride-glucose index on ischemic stroke: a systematic review and meta-analysis. Cardiovasc Diabetol 2023;22:2. [Crossref] [PubMed]
6. Wang X, Feng B, Huang Z, et al. Relationship of cumulative exposure to the triglyceride-glucose index with ischemic stroke: a 9-year prospective study in the Kailuan cohort. Cardiovasc Diabetol 2022;21:66. [Crossref] [PubMed]
7. Wu L, Zhu J, Li C, et al. Association of triglyceride-glucose index with ischemic stroke recurrence in nondiabetic patients with small vessel occlusion: a multicenter hospital-based prospective cohort study. Cardiovasc Diabetol 2022;21:250. [Crossref] [PubMed]
8. Yang Y, Li S, Ren Q, et al. The interaction between triglyceride-glucose index and visceral adiposity in cardiovascular disease risk: findings from a nationwide Chinese cohort. Cardiovasc Diabetol 2024;23:427. [Crossref] [PubMed]
9. Lin H, Jia X, Yin Y, et al. Association of body roundness index with cardiovascular disease and all-cause mortality among Chinese adults. Diabetes Obes Metab 2025;27:2698-707. [Crossref] [PubMed]
10. Yang M, Liu J, Shen Q, et al. Body Roundness Index Trajectories and the Incidence of Cardiovascular Disease: Evidence From the China Health and Retirement Longitudinal Study. J Am Heart Assoc 2024;13:e034768. [Crossref] [PubMed]
11. Schweitzer K. Could the Body Roundness Index One Day Replace the BMI? JAMA 2024;332:1317-8. [Crossref] [PubMed]
12. Qiu L, Xiao Z, Fan B, et al. Association of body roundness index with diabetes and prediabetes in US adults from NHANES 2007-2018: a cross-sectional study. Lipids Health Dis 2024;23:252. [Crossref] [PubMed]
13. Huo RR, Liao Q, Zhai L, et al. Interacting and joint effects of triglyceride-glucose index (TyG) and body mass index on stroke risk and the mediating role of TyG in middle-aged and older Chinese adults: a nationwide prospective cohort study. Cardiovasc Diabetol 2024;23:30. [Crossref] [PubMed]
14. He HM, Xie YY, Chen Q, et al. The synergistic effect of the triglyceride-glucose index and a body shape index on cardiovascular mortality: the construction of a novel cardiovascular risk marker. Cardiovasc Diabetol 2025;24:69. [Crossref] [PubMed]
15. Zhang R, Hong J, Wu Y, et al. Joint association of triglyceride glucose index (TyG) and a body shape index (ABSI) with stroke incidence: a nationwide prospective cohort study. Cardiovasc Diabetol 2025;24:7. [Crossref] [PubMed]
16. Cui C, Qi Y, Song J, et al. Comparison of triglyceride glucose index and modified triglyceride glucose indices in prediction of cardiovascular diseases in middle aged and older Chinese adults. Cardiovasc Diabetol 2024;23:185. [Crossref] [PubMed]
17. Zhou Z, Liu Q, Zheng M, et al. Comparative study on the predictive value of TG/HDL-C, TyG and TyG-BMI indices for 5-year mortality in critically ill patients with chronic heart failure: a retrospective study. Cardiovasc Diabetol 2024;23:213. [Crossref] [PubMed]
18. Wang B, Li L, Tang Y, et al. Joint association of triglyceride glucose index (TyG) and body roundness index (BRI) with stroke incidence: a national cohort study. Cardiovasc Diabetol 2025;24:164. [Crossref] [PubMed]
19. Benn M, Emanuelsson F, Tybjærg-Hansen A, et al. Impact of high glucose levels and glucose lowering on risk of ischaemic stroke: a Mendelian randomisation study and meta-analysis. Diabetologia 2021;64:1492-503. [Crossref] [PubMed]
20. Sacco S, Foschi M, Ornello R, et al. Prevention and treatment of ischaemic and haemorrhagic stroke in people with diabetes mellitus: a focus on glucose control and comorbidities. Diabetologia 2024;67:1192-205. [Crossref] [PubMed]
21. Chen J, Wu Q, Liu H, et al. Predictive value of remnant cholesterol inflammatory index for stroke risk: Evidence from the China health and Retirement Longitudinal study. J Adv Res 2025;76:543-52. [Crossref] [PubMed]
22. Zhao Y, Hu Y, Smith JP, et al. Cohort profile: the China Health and Retirement Longitudinal Study (CHARLS). Int J Epidemiol 2014;43:61-8. [Crossref] [PubMed]
23. Ding C, Ng Fat L, Britton A, et al. Binge-pattern alcohol consumption and genetic risk as determinants of alcohol-related liver disease. Nat Commun 2023;14:8041. [Crossref] [PubMed]
24. de Hond AAH, Steyerberg EW, van Calster B. Interpreting area under the receiver operating characteristic curve. Lancet Digit Health 2022;4:e853-5. [Crossref] [PubMed]
25. Lu R, Lumish HS, Hasegawa K, et al. Prediction of new-onset atrial fibrillation in patients with hypertrophic cardiomyopathy using machine learning. Eur J Heart Fail 2025;27:275-84. [Crossref] [PubMed]
26. Wu Y, Yang Y, Zhang J, et al. The change of triglyceride-glucose index may predict incidence of stroke in the general population over 45 years old. Cardiovasc Diabetol 2023;22:132. [Crossref] [PubMed]
27. Chen Y, Xie K, Han Y, et al. The association between triglyceride-glucose index and its combination with systemic inflammation indicators and all-cause and cardiovascular mortality in the general US population: NHANES 1999-2018. Lipids Health Dis 2024;23:289. [Crossref] [PubMed]
28. Dui X, Chen X, Zhu L, et al. The association between derived TyG index and the risk of heart failure in the elderly population: a prospective cohort study from 2017 to 2023. BMC Public Health 2025;25:863. [Crossref] [PubMed]
29. Tang X, Zhang K, He R. The association of triglyceride-glucose and triglyceride-glucose related indices with the risk of heart disease in a national. Cardiovasc Diabetol 2025;24:54. [Crossref] [PubMed]
30. Yue Y, Li P, Sun Z, et al. Association of novel triglyceride-glucose-related indices with incident stroke in early-stage cardiovascular-kidney-metabolic syndrome. Cardiovasc Diabetol 2025;24:301. [Crossref] [PubMed]
31. Xu J, Cai D, Jiao Y, et al. Insights into the complex relationship between triglyceride glucose-waist height ratio index, mean arterial pressure, and cardiovascular disease: a nationwide prospective cohort study. Cardiovasc Diabetol 2025;24:93. [Crossref] [PubMed]
32. Yao F, Cui J, Shen Y, et al. Evaluating a new obesity indicator for stroke risk prediction: comparative cohort analysis in rural settings of two nations. BMC Public Health 2024;24:3301. [Crossref] [PubMed]
33. Chen X, Du X, Lu F, et al. The Association Between the Triglyceride-Glucose Index, Its Combination with the Body Roundness Index, and Chronic Kidney Disease in Patients with Type 2 Diabetes in Eastern China: A Preliminary Study. Nutrients 2025;17:492. [Crossref] [PubMed]
34. Wang L, Liu J, Wang J, et al. Joint association of triglyceride glucose index (TyG) and body roundness index with the risk of periodontitis: a cross-sectional study. Front Nutr 2025;12:1642112. [Crossref] [PubMed]
35. Wu M, Yu X, Xu L, et al. Associations of longitudinal trajectories in body roundness index with mortality and cardiovascular outcomes: a cohort study. Am J Clin Nutr 2022;115:671-8. [Crossref] [PubMed]
36. Sun Y, Ji H, Sun W, et al. Triglyceride glucose (TyG) index: A promising biomarker for diagnosis and treatment of different diseases. Eur J Intern Med 2025;131:3-14. [Crossref] [PubMed]
37. Hill MA, Yang Y, Zhang L, et al. Insulin resistance, cardiovascular stiffening and cardiovascular disease. Metabolism 2021;119:154766. [Crossref] [PubMed]
38. Koenen M, Hill MA, Cohen P, et al. Obesity, Adipose Tissue and Vascular Dysfunction. Circ Res 2021;128:951-68. [Crossref] [PubMed]
39. Dang K, Wang X, Hu J, et al. The association between triglyceride-glucose index and its combination with obesity indicators and cardiovascular disease: NHANES 2003-2018. Cardiovasc Diabetol 2024;23:8. [Crossref] [PubMed]
40. Wang M, Gao B, Huang F. Association between the triglyceride glucose-Chinese visceral adiposity index and new-onset stroke risk: a national cohort study. Cardiovasc Diabetol 2025;24:119. [Crossref] [PubMed]
41. Scurt FG, Ganz MJ, Herzog C, et al. Association of metabolic syndrome and chronic kidney disease. Obes Rev 2024;25:e13649. [Crossref] [PubMed]