Relationship Between a Novel Model of Insulin Sensitivity and Arterial

Introduction

The prevalence of diabetes, particularly type 2 diabetes (T2D), is rising globally, posing a significant public health challenge due to its various acute and chronic complications.¹ Among these complications, cardiovascular diseases (CVDs) stand out as the leading cause of death in patients with T2D.² Vascular dysfunctions, including arterial stiffness (AS) and impaired vasodilation, can emerge before the onset of severe CVDs symptoms.³ Therefore, early assessment of AS is particularly important in the management of T2D. The brachial-ankle pulse wave velocity (baPWV) is a simple, effective and non-invasive method for evaluating AS,⁴ and can independently predict cardiovascular risk, providing important evidence for assessing the development of CVDs in individuals.⁵

Insulin resistance (IR) is considered a significant factor to AS and the progression of CVDs.⁶ While the euglycemic hyperinsulinemic clamp (EHC) is considered the gold standard for assessing IR,⁷ its invasive nature, time consuming and requirement for hospitalization limit its practical applicability. The homeostasis model assessment index (HOMA-IR) offers a simpler approach to assessing IR.⁸ However, it presents specific challenges for patients undergoing insulin therapy. Recently, a growing number of non-insulin-based IR surrogate markers have been proposed, including the triglyceride-glucose (TyG) index, triglyceride-to-high-density lipoprotein cholesterol (TG/HDL-c) ratio, and metabolic score for insulin resistance (METS-IR).^9–11 These markers have been associated with various metabolic diseases. One of our previous studies examined their relationship with nonalcoholic fatty liver disease (NAFLD) in patients with T2D, highlighting their clinical relevance in this context.¹² Building on this foundation, our current study shifts the focus toward AS, a distinct yet critical cardiovascular complication in T2D. A more recent development is the natural log transformation of the glucose disposal rate (log_e GDR), a non-insulin-based model to assess insulin sensitivity (IS) in T2D patients.¹³ The log_e GDR is calculated based on body mass index (BMI), triglycerides (TG), the urinary albumin to creatinine ratio (UACR) and γ-glutamyl transferase (GGT), and it has been validated and demonstrated a strong association with CVDs and mortality rates.¹³ Despite these advances, no studies to date have specifically investigated the relationship between the log_e GDR and AS in T2D. Based on the groundwork laid by our earlier research, this study aims to explore the novel association between log_e GDR and AS, offering fresh insights into the complex interplay between IS and vascular health.

The prevalence of non-obese T2D is gradually increasing, particularly in Asian countries.^14,15 Although CVDs and other conditions have traditionally been associated with obesity and being overweight, recent evidence suggests that non-obese T2D patients may have higher all-cause and cardiovascular mortality rates.^16,17 This could be attributed to factors such as increased visceral fat, impaired IS, and heightened inflammatory responses despite a normal BMI.¹⁸ Furthermore, recent studies have reported non-obese T2D patients have a comparable or even higher prevalence of AS compared to their obese counterparts¹⁸ Due to significant differences in metabolic characteristics between obese and non-obese diabetic patients, particularly regarding IS.¹⁹ And research exploring the relationship between IR surrogate markers and AS in non-obese patients with T2D remains limited. Therefore, we aim to analyze the relationship between the log_e GDR and AS in this population.

Methods

Study Design and Population

We retrospectively reviewed the medical records of patients aged ≥ 18 years with T2D from the Department Endocrinology of Linyi People’s Hospital, from January 2020 to March 2023. The exclusion criteria were (1) subjects missing key anthropometric measurements (height and weight); (2) subjects who had severe liver and kidney dysfunctions; (3) subjects with a history of angina, myocardial infarction and cerebrovascular accident; (4) subjects who had not undergone the baPWV tests or whose clinical data were incomplete, including GGT, TG and UACR; (5) subjects with BMI ≥ 24 kg/m². Ultimately, a total of 790 non-obese patients with T2D were eventually included (Figure 1).

Figure 1 The flow chart of study participants selection.

It is important to note that a portion of the participants included in the current study were also part of our previous work, which primarily investigated the association between IR markers and NAFLD in the overall T2D population. In terms of exclusion criteria, the previous study mainly excluded confounding factors that could affect the diagnosis and analysis of fatty liver disease.¹² For the overlapping populations, we further compared differences in baseline characteristics and various IR indices between the two studies, and the results did not show significant differences, suggesting that the results of this study are relatively stable and have some replication.

Demographic Information

The sex, age, diabetes duration and self-reported current cigarette smoking and drinking status were collected.

Physical Examinations

According to unified standards, the height, weight, systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured and collected. The bioelectrical impedance analysis (Omron DUALSCAN HDS-2000, Kyoto, Japan) was used to measure the visceral fat area (VFA) and subcutaneous fat area (SFA).

Each participant’s baPWV was measured using the automated system BP-203RPE III (Omron Healthcare Co., Japan) by trained technicians. The device simultaneously recorded pulse waveforms from the brachial and tibial arteries and automatically calculated baPWV values. Before measurement, participants were required to rest in a supine position for at least 5 minutes to ensure hemodynamic stability. Subsequently, appropriately sized cuffs were placed on both upper arms and ankles, and the device was operated according to standard protocols to obtain waveform signals and compute baPWV values.²⁰ To enhance measurement accuracy, this study analyzed data in cases where there was a significant difference between left and right baPWV values and assessed each side’s baPWV separately. AS was defined as baPWV ≥1800 cm/s.

Laboratory Measurements

Following an overnight fast, blood samples were collected and analyzed in the morning for alanine aminotransferase (ALT), aspartate aminotransferase (AST), GGT, TG, HDL-c, total cholesterol (TC), low density lipoprotein-cholesterol (LDL-c), serum creatinine (Scr), uric acid (UA), Cystatin C (Cys C), hemoglobin (Hb), fasting blood glucose (FBG) and glycosylated haemoglobin (HbA1c), fasting insulin (FINS) and UACR. A comprehensive overview of the tools and methods utilized in this research is available in our earlier publication.¹² Non-obese was defined as BMI < 24 kg/m².

Parameter Calculation

1. BMI = weight (kg) / height² (m²);
2. eGFR = 175 * Scr (mg/dL) ^−1.234 * age ^−0.179 * (0.79, if female);²¹
3. HOMA-IR = FBG (mmol/L) * FINS (µU/mL)/22.5;⁸
4. TyG index = ln [TG (mg/dL) × FBG (mg/dL)/2];¹⁰
5. TG/HDL-c ratio = TG (mmol/L)/HDL-c (mmol/L);¹¹
6. METS-IR = ln [(2*FBG (mg/dL)) + TG (mg/dL)] *BMI)/(Ln [HDL-c (mg/dL)]);⁹
7. Log_e GDR = 5.3505–0.3697 * log_e (GGT, IU/L) – 0.2591 * log_e (TG, mg/dL) – 0.1169 * log_e (UACR, mg/g) – (0.0279*BMI, kg/m²).¹³

Statistical Analysis

Statistical analysis was performed using SPSS 26.0 (SPSS Inc, Chicago, USA) and R (version 4.3.2). Data were presented as means ± SD for normally distributed variables and as medians (interquartile ranges) for non-normally distributed variables. Independent-Samples T test and Mann–Whitney U-test were used for comparisons of normally and abnormally distributed continuous variables between two groups, respectively. Categorical variables were presented as percentage (%) and were compared by Chi-square test. For normally distributed data, an Analysis of Variance (ANOVA) and Student-Newman-Keuls tests were used for multiple and pairwise comparisons between the log_e GDR tertiles groups, while the Kruskal–Wallis one-way ANOVA test was used for abnormally distributed data. Pearson correlation and multiple linear stepwise regression analyses were used to evaluate the independent correlations of baPWV. Univariate logistic regression analysis and directed acyclic graphs (DAG) were used to guide the selection of covariates for AS. The DAG was constructed using the dagitty package. And the identified minimal adjustment set includes age, BMI, diabetes duration, FBG, TG, HOMA-IR, METS-IR, TG/HDL-c ratio and TyG index. Logistic regression analysis was used to analyze the independent correlates of AS. Net reclassification improvement (NRI) analysis was performed using the survIDINRI package in R to assess the incremental predictive value of log_e GDR compared with other IR markers for identifying AS. Statistical differences were defined by P-value (two-tailed) less than 0.05.

Results

Clinical and Biochemical Characteristics

The clinical and biochemical characteristics of the participants are shown in Table 1. A total of 790 non-obese patients with T2D were enrolled in our study. The subjects were divided into two groups including non-AS group (baPWV < 1800cm/s) and AS group (baPWV ≥ 1800cm/s). Compared with the non-AS group, the age, diabetes duration, VFA, SFA, SBP, DBP, AST, GGT, UA, Scr, UACR and Cys C were increased in AS group, but the HbA1c, eGFR, Hb and log_e GDR were markedly reduced (all P < 0.05). There were no obvious differences in BMI, TC, LDL-c, HDL-c, TG, FBG, FINS, ALT, HOMA-IR, TG/HDL-c ratio, TyG index, METS-IR and the percentages of males, smoking and drinking between the two groups (all P > 0.05).

Table 1 Clinical and Biochemical Characteristics by Presence of AS

Then, according to tertiles of log_e GDR, the participants were divided into three groups: T1 (0.25–1.98), T2 (1.98–2.28) and T3 (2.28–3.12) (Table 2). As the log_e GDR tertiles increased, the age, diabetes duration, BMI, VFA, SFA, SBP, DBP, TC, LDL-c, TG, FINS, HbA1c, ALT, AST, GGT, UA, Scr, UACR, Cys C, HOMA-IR, TyG index, TG/HDL-c ratio, METS-IR, baPWV, the percentages of smoking, drinking and AS were gradually decreased, while the HDL-c, Hb and eGFR were gradually elevated (all P < 0.05). The FBG and the percentages of males were no significant different between the three groups (both P > 0.05).

Table 2 Comparison of Variables According to the Tertiles of Loge GDR

Correlation Between baPWV or AS and Each Variable by Univariate Analysis

As shown in Table 3, a Pearson correlation analysis was performed to analyze the association between baPWV and each variable. The results displayed that the baPWV was positively related to age, diabetes duration, VFA, SFA, SBP, DBP, TG, FINS, GGT, UA, UACR, Cys C and TG/HDL-c ratio, while negatively to the Hb, HbA1c, eGFR and log_e GDR (all P < 0.05). BMI, TC, LDL-c, HDL-c, FBG, ALT, AST, HOMA-IR, TyG index and METS-IR were not correlated with baPWV (all P > 0.05).

Table 3 The Correlation Between baPWV or AS and Different Variables by Univariate Analysis

Moreover, univariate regression analysis was conducted to identify the factors associated with AS. The results showed that AS was positively related to the age, diabetes duration, VFA, SFA, SBP, DBP, FINS, AST, UA, UACR, Cys C, and negatively to the Hb, HbA1c, eGFR and log_e GDR (all P < 0.05). No significant relationships existed between AS and BMI, TC, LDL-c, TG, HDL-c, FBG, ALT, GGT, HOMA-IR, TyG index, TG/HDL-c ratio, METS-IR and the percentages of males, smoking and drinking (all P > 0.05).

Independent Variables of baPWV by Multiple Linear Stepwise Regression Analysis

The covariates for multivariate linear regression analysis were determined based on the results of Pearson correlation analysis and previous literature reports. A multiple linear stepwise regression analysis was conducted to analyze the independent correlations of baPWV (Table 4). The age, diabetes duration, VFA, SFA, SBP, DBP, TG, FINS, GGT, UA, UACR, Cys C, TG/HDL-c ratio, Hb, HbA1c, eGFR and log_e GDR were set as the dependent variables based on the results of Pearson correlation analysis, and the results displayed that the age, SBP and log_e GDR fit a regression model (all P < 0.05).

Table 4 Multivariate Linear Regression Analysis with baPWV as the Dependent Variable

Independent Correlations of AS by Logistic Regression Analysis

Finally, AS was served as the dependent variable, and based on the results of univariate logistic regression analysis, the DAG diagram (Figure 2), and previous literature, the following variables were included as independent variables: age, diabetes duration, VFA, SFA, SBP, DBP, FINS, HbA1c, AST, UA, eGFR, UACR, Cys C, Hb, BMI, FBG, TG, HOMA-IR, TG/HDL-c ratio, TyG index, METS-IR, log_e GDR and the percentages of smoking and drinking. A logistic regression analysis was performed to analyze the independent correlates of AS (Table 5), and the results found that after adjusting for the other variables, the log_e GDR (OR: 0.286, 95.0% CI for OR: 0.110–0.743), age (OR: 1.196, 95.0% CI for OR: 1.138–1.258), SBP (OR: 1.053, 95.0% CI for OR: 1.031–1.075) and FBG (OR: 0.886, 95.0% CI for OR: 0.792–0.990) were independently related to AS.

Table 5 The Independent Variables for AS

Figure 2 The DAG of identifying confounding variables.

Predictive Value of IR Markers for AS

To assess the incremental predictive value of various IR markers for AS, NRI analysis was performed based on logistic regression models (Table 6). All models were adjusted for potential confounders, including age, diabetes duration, VFA, SFA, SBP, DBP, FINS, HbA1c, AST, UA, eGFR, UACR, Cys C, Hb, BMI, FBG, TG, smoking and drinking. Building upon the base model without any IR marker, integrating log_e GDR yielded a modest improvement in the model’s ability to reclassify patients with AS (NRI:0.043, 95% CI 0.009–0.079, P = 0.011). In contrast, building upon the base model, integrating other IR markers such as HOMA-IR (NRI:0.007, P = 0.697), TyG index (NRI:0.011, P = 0.356), TG/HDL-c ratio (NRI:0.006, P = 0.317), and METS-IR (NRI: −0.004, P = 0.568) did not significantly improve the predictive performance.

Table 6 Analysis of the NRI for Predicting AS

Discussion

This cross-sectional study of non-obese patients with T2D revealed a significant negative association between the log_e GDR and both baPWV and AS. Increased log_e GDR tertiles corresponded with a significant reduction in baPWV and AS incidence. Furthermore, after adjusting for confounding factors, the log_e GDR was independently associated with baPWV and AS.

IR is common among diabetic patients, leading to endothelial dysfunction and inflammatory responses that contribute to AS and atherosclerosis.²² Although the EHC is considered the gold standard for assessing IS, its complexity, time consuming, and requirement for specialized personnel limit its use in large-scale clinical studies. HOMA-IR is a commonly used and simpler indicator of IR, but it relies on FINS. Previous studies have shown that fluctuations in insulin levels can be significantly influenced by an individual’s glucose tolerance and the effects of treatment. Therefore, FINS levels may not be entirely accurate for patients with T2D undergoing treatment.^23,24 Recently, an increasing number of studies have explored the close association between non-insulin-based IR surrogate indicators and AS across various populations. For instance, a study in a healthy Japanese cohort found a significant correlation between the METS-IR and AS.²⁵ A study involving 1895 participants showed a close correlation between the TyG index and the TG/HDL-c ratio with AS in hypertensive patients, while no such relationship was observed in patients with prehypertension.²⁶ Furthermore, research on patients with T2D had indicated that the TyG index was independently and more strongly associated with the prevalence of increased AS compared to HOMA-IR.²⁰ The relationship between non-insulin-based IR surrogate indicators with AS had also been validated in lean postmenopausal women, Chinese non-hypertensive and older subjects.^27–29

The log_e GDR is a newly developed model for assessing IS in T2D, and it has been validated as a reliable EHC-based surrogate capable of capturing the variability of IS in patients with T2D well.¹³ The inclusion of metabolic components (GGT, UACR, BMI and TG) allows log_e GDR to reflect a more comprehensive metabolic profile and potentially capturing a broader range of pathogenic mechanisms. In our study, we found that it was closely associated with IR markers as well. As the tertiles of log_e GDR increased, significant reductions were observed in other IR markers, suggesting a consistent relationship between log_e GDR and IS. Notably, we found that the log_e GDR was independently related to baPWV and AS. This relationship remains important even after adjusting for other confounding factors including IR markers (HOMA-IR, TG/HDL-c ratio, TyG index, and METS-IR).

The mechanisms potentially linking log_e GDR to AS are likely multifactorial and may involve several key pathways. The components included in the calculation of log_e GDR, including GGT, UACR, BMI and TG, may have been suggested as part of circadian syndrome.³⁰ Recent studies indicate that circadian syndrome may be a better predictor of CVDs risk than metabolic syndrome,³⁰ suggesting that log_e GDR might reflect a disruption in circadian rhythms, potentially influencing cardiovascular health. GGT is a key marker of oxidative stress, promoting endothelial dysfunction by reducing nitric oxide bioavailability and increasing vascular inflammation, both of which contribute to arterial stiffening. TG facilitates lipid accumulation in the vascular wall, leading to foam cell formation and atherosclerosis progression. Elevated TG levels are also associated with increased production of small, dense LDL particles, which enhance oxidative stress and vascular inflammation. UACR reflects endothelial dysfunction and vascular damage, as albuminuria is linked to increased vascular permeability and low-grade inflammation, both contributing to arterial remodeling. Additionally, BMI, particularly in the context of visceral adiposity, is associated with chronic low-grade inflammation and activation of the renin-angiotensin-aldosterone system, further promoting vascular stiffness. These components effectively represent the key metabolic pathways leading to AS, supporting the close relationship between log_e GDR and AS.^31–33

Additionally, AS is a degenerative vascular process that increases with age.³⁴ High SBP levels may damage endothelial function, leading to progressively stiffer arteries.³⁵ Be consistent with the above findings, we found a strong relationship between age and SBP with AS in non-obese patients with T2D. This underscores the importance of managing SBP as a modifiable risk factor for AS, particularly in this population. Interestingly, we observed a negative correlation between AS and FBG, which was inconsistent with most studies that suggested elevated FBG was a significant risk factor for AS.³⁶ The multifaceted influencing factors of AS may help explain this phenomenon. As mentioned earlier, the average age in the AS group was significantly higher than that in the non-AS group, and some studies have suggested that older diabetic patients tend to have better blood glucose control.³⁷

The relationship between the novel IS index log_e GDR and AS has not been extensively studied in the context of non-obese T2D. Our study is the first to observe a strong association between log_e GDR and AS in non-obese patients with T2D, highlighting its potential clinical significance. Although non-obese individuals with T2D may have normal body weight, they can still exhibit significant vascular changes. Since log_e GDR incorporates metabolic parameters including BMI, TG, UACR and GGT, it may reflect a broader metabolic disorder amenable to intervention than other IR markers. Importantly, logₑ GDR demonstrated the highest NRI among the evaluated IR indicators, indicating relatively better discriminatory capacity for AS. However, the overall improvement in risk prediction was modest, suggesting that its incremental value in risk stratification may be limited. Therefore, while logₑ GDR shows potential as a complementary tool for early identification of cardiovascular risk in non-obese T2D patients, its clinical utility should be interpreted with caution. Further prospective studies with larger, diverse cohorts are needed to confirm these findings and to clarify the role of logₑ GDR in improving cardiovascular risk prediction models.

Several limitations of this study should be acknowledged. First, as with all cross-sectional studies, we cannot establish causality between log_e GDR and AS. Longitudinal studies are essential to determine the temporal relationship and causal pathways between these variables. Second, using BMI < 24 kg/m² to define “non-obese” may not perfectly capture individuals with increased visceral adiposity, which is a key driver of metabolic dysfunction. Future studies could consider including measures such as waist circumference or waist-to-hip ratio, which provide more direct insight into visceral fat distribution. Lastly, this study is single-center and based on a small sample size, which may limit the generalizability of the results. Future prospective multi-center studies involving larger populations are needed to confirm these findings and further investigate the underlying mechanisms.

Conclusion

In conclusion, the log_e GDR, as a new simple index of IS, is independently associated with AS in non-obese patients with T2D. Its inclusion in existing risk models modestly improved the identification of arterial stiffness. The potential utility of log_e GDR in cardiovascular risk assessment warrants further investigation and validation in future studies.

Ethics Approval and Consent to Participate

The study was approved by the Human Ethics Committee of the Linyi People’s Hospital. All procedures were performed in accordance with ethical standards laid out in the Declaration of Helsinki. Informed consent was obtained from the patients.

Acknowledgments

Shuwei Shi is currently Department of Endocrinology, Linyi People’s Hospital Affiliated to Shandong Second Medical University, Linyi, China. This study was conducted while she was affiliated with the School of Clinical Medicine, Shandong Second Medical University, Weifang, Shandong, China. Baolan Ji and Guanqi Gao are co-corresponding authors for this study.

Funding

This study was supported by grants from the Postdoctoral Program of Affiliated Hospital of Jining Medical University (JYFY322152).

Disclosure

All authors declare that they have no competing interests in this study.

References

1. GBD. 2021 Diabetes Collaborators. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet. 2023;402(10397):203–234. doi:10.1016/S0140-6736(23)01301-6.

2. Cx M, Xn M, Guan CH, Li YD, Mauricio D, Fu SB. Cardiovascular disease in type 2 diabetes mellitus: progress toward personalized management. Cardiovascular Diabetol. 2022;21(1):74. doi:10.1186/s12933-022-01516-6

3. Cai L, Shen W, Li J, et al. Association between glycemia risk index and arterial stiffness in type 2 diabetes. J Diabetes Invest. 2024;15(5):614. doi:10.1111/jdi.14153

4. Munakata M. Brachial-ankle pulse wave velocity in the measurement of arterial stiffness: recent evidence and clinical applications. Curr Hypertens Rev. 2014;10(1):49–57. doi:10.2174/157340211001141111160957

5. Chirinos JA, Segers P, Hughes T, Townsend R. Large Artery Stiffness in Health and Disease: JACC State-of-the-Art Review. J Am College Cardiol. 2019;74(9):1237. doi:10.1016/j.jacc.2019.07.012

6. Hill MA, Yang Y, Zhang L, et al. Insulin resistance, cardiovascular stiffening and cardiovascular disease. Metabolism. 2021;119:154766. doi:10.1016/j.metabol.2021.154766

7. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979;237(3):E214–223. doi:10.1152/ajpendo.1979.237.3.E214

8. Dr M, Jp H, As R, Ba N, Df T, Rc T. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):1. doi:10.1007/BF00280883

9. Bello-Chavolla OY, Almeda-Valdes P, Gomez-Velasco D, et al. METS-IR, a novel score to evaluate insulin sensitivity, is predictive of visceral adiposity and incident type 2 diabetes. Eur J Endocrinol. 2018;178(5):533–544. doi:10.1530/EJE-17-0883

10. Simental-Mendía LE, Rodríguez-Morán M, Guerrero-Romero F. The product of fasting glucose and triglycerides as surrogate for identifying insulin resistance in apparently healthy subjects. Metab Syndr Relat Disord. 2008;6(4):299–304. doi:10.1089/met.2008.0034

11. Giannini C, Santoro N, Caprio S, et al. The triglyceride-to-HDL cholesterol ratio: association with insulin resistance in obese youths of different ethnic backgrounds. Diabetes Care. 2011;34(8):1869–1874. doi:10.2337/dc10-2234

12. Ma X, Ji B, Du W, et al. METS-IR, a Novel Simple Insulin Resistance Index, is Associated with NAFLD in Patients with Type 2 Diabetes Mellitus. Diabetes Metab Syndr Obes. 2024;17:3481–3490. doi:10.2147/DMSO.S476398

13. Ciardullo S, Dodesini AR, Lepore G, et al. Development of a New Model of Insulin Sensitivity in Patients With Type 2 Diabetes and Association With Mortality. J Clin Endocrinol Metab. 2024;109(5):1308–1317. doi:10.1210/clinem/dgad682

14. Yu HJ, Ho M, Liu X, Yang J, Chau PH, Fong DYT. Incidence and temporal trends in type 2 diabetes by weight status: a systematic review and meta-analysis of prospective cohort studies. J Glob Health. 2023;13:04088. doi:10.7189/jogh.13.04088

15. Adesoba TP, Brown CC. Trends in the Prevalence of Lean Diabetes Among U.S. Adults, 2015–2020. Diabetes Care. 2023;46(4):885–889. doi:10.2337/dc22-1847

16. Carnethon MR, De Chavez PJD, Biggs ML, et al. Association of weight status with mortality in adults with incident diabetes. JAMA. 2012;308(6):581–590. doi:10.1001/jama.2012.9282

17. Flegal KM, Kit BK, Orpana H, Graubard BI. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. JAMA. 2013;309(1):71–82. doi:10.1001/jama.2012.113905

18. Bouchi R, Minami I, Ohara N, et al. Impact of increased visceral adiposity with normal weight on the progression of arterial stiffness in Japanese patients with type 2 diabetes. BMJ Open Diabetes Res Care. 2015;3(1):e000081. doi:10.1136/bmjdrc-2015-000081

19. Gudipaty L, Rosenfeld NK, Fuller CS, Cuchel M, Rickels MR. Different β-cell secretory phenotype in non-obese compared to obese early type 2 diabetes. Diabetes/Metab Res Rev. 2020;36(5):e3295. doi:10.1002/dmrr.3295

20. Wang S, Shi J, Peng Y, et al. Stronger association of triglyceride glucose index than the HOMA-IR with arterial stiffness in patients with type 2 diabetes: a real-world single-centre study. Cardiovasc Diabetol. 2021;20(1):82. doi:10.1186/s12933-021-01274-x

21. An L, Yu Q, Chen L, et al. The Association Between the Decline of eGFR and a Reduction of Hemoglobin A1c in Type 2 Diabetic Patients. Front Endocrinol (Lausanne). 2021;12:723720. doi:10.3389/fendo.2021.723720

22. Tan J, Li X, Dou N. Insulin Resistance Triggers Atherosclerosis: caveolin 1 Cooperates with PKCzeta to Block Insulin Signaling in Vascular Endothelial Cells. Cardiovasc Drugs Ther. 2023;38(5):885. doi:10.1007/s10557-023-07477-6

23. Liang L, fen FJ, Chun ZC, Hong F, Wang C-L, Wang X-M. Wang C lin, Wang X min. [Metformin hydrochloride ameliorates adiponectin levels and insulin sensitivity in adolescents with metabolic syndrome]. Zhonghua Er Ke Za Zhi. 2006;44(2):118–121.

24. Jayagopal V, Kilpatrick ES, Jennings PE, Hepburn DA, Atkin SL. Biological variation of homeostasis model assessment-derived insulin resistance in type 2 diabetes. Diabetes Care. 2002;25(11):2022–2025. doi:10.2337/diacare.25.11.2022

25. Liu G. Association between the metabolic score for insulin resistance (METS-IR) and arterial stiffness among health check-up population in Japan: a retrospective cross-sectional study. Front Endocrinol. 2024;14:1308719. doi:10.3389/fendo.2023.1308719

26. Wu Z, Zhou D, Liu Y, et al. Association of TyG index and TG/HDL-C ratio with arterial stiffness progression in a non-normotensive population. Cardiovascular Diabetol. 2021;20(1):134. doi:10.1186/s12933-021-01330-6

27. Su Y, Wang S, Sun J, et al. Triglyceride Glucose Index Associated With Arterial Stiffness in Chinese Community-Dwelling Elderly. Front Cardiovasc Med. 2021;8:737899. doi:10.3389/fcvm.2021.737899

28. Lambrinoudaki I, Kazani MV, Armeni E, et al. The TyG Index as a Marker of Subclinical Atherosclerosis and Arterial Stiffness in Lean and Overweight Postmenopausal Women. Heart Lung Circ. 2018;27(6):716–724. doi:10.1016/j.hlc.2017.05.142

29. Zhang X, Ye R, Yu C, Liu T, Chen X. Correlation Between Non-insulin-Based Insulin Resistance Indices and Increased Arterial Stiffness Measured by the Cardio-Ankle Vascular Index in Non-hypertensive Chinese Subjects: a Cross-Sectional Study. Front Cardiovasc Med. 2022;9:903307. doi:10.3389/fcvm.2022.903307

30. Shi Z, Tuomilehto J, Kronfeld-Schor N, et al. The circadian syndrome predicts cardiovascular disease better than metabolic syndrome in Chinese adults. J Intern Med. 2021;289(6):851–860. doi:10.1111/joim.13204

31. Wildman RP, Mackey RH, Bostom A, Thompson T, Sutton-Tyrrell K. Measures of obesity are associated with vascular stiffness in young and older adults. Hypertension. 2003;42(4):468–473. doi:10.1161/01.HYP.0000090360.78539.CD

32. Stehouwer CDA, Smulders YM. Microalbuminuria and risk for cardiovascular disease: analysis of potential mechanisms. J Am Soc Nephrol. 2006;17(8):2106–2111. doi:10.1681/ASN.2005121288

33. Lee DH, Jacobs DRJ. Serum gamma-glutamyltransferase: new insights about an old enzyme. J Epidemiol Community Health. 2009;63(11):884–886. doi:10.1136/jech.2008.083592

34. Lu Y, Kiechl SJ, Wang J, et al. Global distributions of age- and sex-related arterial stiffness: systematic review and meta-analysis of 167 studies with 509,743 participants. EBioMed. 2023;92:104619. doi:10.1016/j.ebiom.2023.104619

35. Liu R, Li D, Yang Y, Hu Y, Wu S, Tian Y. Systolic Blood Pressure Trajectories and the Progression of Arterial Stiffness in Chinese Adults. Int J Environ Res Public Health. 2022;19(16):10046. doi:10.3390/ijerph191610046

36. Fu S, Chen W, Luo L, Ye P. Roles of fasting and postprandial blood glucose in the effect of type 2 diabetes on central arterial stiffness: a 5-year prospective community-based analysis. Diabetol Metab Syndr. 2017;9(1):33. doi:10.1186/s13098-017-0231-3

37. Shamshirgaran SM, Mamaghanian A, Aliasgarzadeh A, Aiminisani N, Iranparvar-Alamdari M, Ataie J. Age differences in diabetes-related complications and glycemic control. BMC Endocr Disord. 2017;17(1):25. doi:10.1186/s12902-017-0175-5