Association between the triglyceride-glucose index and vascular damage among community residents in Yunnan, China Authors: Wenlong Zhu, Nan Zhang, Yu Xia, Rui Zhu, Wengang Shi, Lin Duo, Mingjing Tang, Da Zhu Published: 2025 ABSTRACT BackgroundThe triglyceride-glucose (TyG) index is a straightforward and widely used surrogate marker for identifying insulin resistance (IR). The association between the TyG index and both macrovascular and microvascular damage remains unclear. This study aims to investigate the association between vascular damage and IR in community-dwelling adults in Yunnan Province, China.MethodsA total of 3201 participants were involved. Ankle-brachial index (ABI), estimated glomerular filtration rate, and the urine albumin-to-creatinine ratio were measured and calculated. The TyG index was calculated as ln[triglycerides (mg/dL) × fasting blood glucose (mg/dL)/2]. Logistic regression, subgroup and interaction analyses, and restricted cubic spline regression were employed to assess the relationship between the TyG index and vascular damage.ResultsIn the univariate logistic regression model, a 1-unit increase in the TyG index was associated with 60.75%, 59.35%, and 129.60% higher odds of ABI ≤ 0.9 (OR: 1.61, 95% CI: 1.35 to 1.91), increased albuminuria (OR: 1.59, 95% CI: 1.34 to 1.89), and chronic kidney disease (CKD) (OR: 2.30, 95% CI: 1.65 to 3.20), respectively. These associations remained significant after adjustment for demographic, behavioral, and metabolic factors. The relationship between the TyG index and the presence of ABI ≤ 0.9 and increased albuminuria was not influenced by other factors (p for interaction > 0.05). The higher odds of CKD associated with each unit rise in the TyG index was greater in participants with normal weight or normal lipid levels compared to those who were overweight/obese or had dyslipidemia (p for interaction < 0.05). The odds of ABI ≤ 0.9 and CKD increased nonlinearly with increasing TyG index when TyG index was below 8.91, but stabilized when TyG index was exceeded 8.91 (p for nonlinear = 0.033). The likelihood of increased albuminuria increased with an advancing TyG index (p = 0.026), but the relationship was not nonlinear (p for nonlinear = 0.352).ConclusionsAn elevated TyG index was independently associated with an increased odds of macro- and microvascular damage, suggesting that the TyG index may serve as a biomarker for vascular damage. The relationship between CKD and IR was affected by overweight/obesity and dyslipidemia.Trial registrationNot applicable.Clinical trial numberNot applicable.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12902-025-02145-2. Background Cardiovascular disease (CVD) is a leading cause of mortality and morbidity globally, contributing to a substantial burden of disease [1, 2]. The progression of CVD is driven by vascular dysfunction and damage [3, 4]. Previous studies have shown that both macrovascular damage (such as peripheral artery disease (PAD)) and microvascular damage (such as increased albuminuria and chronic kidney disease (CKD)) are significant risk factors for CVD-related mortality and morbidity [3, 5–9]. PAD is characterized by an ankle-brachial index (ABI) of less than 0.90. A meta-analysis estimated the prevalence of PAD in China to be between 2.81% and 3.84% in people aged 25–29 years, rising to between 21.95% and 27.95% in those aged 95–99 years [10]. Additionally, a nationwide survey revealed that 6.6% of adults aged 35 years and older had PAD [11]. The prevalence of CKD in China was estimated at 8.2% in 2019 [12]. Microalbuminuria has been recognized as an early indicator of renal damage and serves as an independent predictive and prognostic marker for CVD [8]. Identifying the risk factors associated with both macro- and microvascular damage is crucial for mitigating the disease burden related to vascular complications and CVD. Insulin resistance (IR) is defined as an experimental or clinical condition in which insulin exhibits a lower biological effect than expected, playing a key role in the development of metabolic syndrome and CVD [13]. IR can be assessed using methods such as the hyperinsulinemic euglycemic clamp (HEC), the homeostasis model assessment of insulin resistance (HOMA-IR), or the triglyceride-glucose (TyG) index [14]. Due to the complexity of the HEC, HOMA-IR and TyG are straightforward and widely adopted approaches for identifying IR in population studies. Many studies have shown a strong correlation between HOMA-IR and the TyG index with HEC, with the TyG index showing superior performance compared to HOMA-IR in identifying IR [15, 16]. Previous studies have indicated that IR was a risk factor for CVD [17–19], heart failure [20], arterial stiffness [21, 22], PAD [23], carotid atherosclerosis [24], CKD [25], microalbuminuria [26], and obstructive sleep apnea [27]. A meta-analysis showed that higher levels of IR were associated with a higher risk of developing heart failure after accounting for traditional risk factors [20]. Follow-up studies of diabetic patients by Bonora et al. and Hsu et al. found that HOMA-IR was an independent predictor of CVD and microalbuminuria in individuals with type 2 diabetes [18, 26]. A study by Hanley et al. involving Mexican-American and non-Hispanic white residents revealed a significant association between HOMA-IR and risk of CVD after adjustment for multiple covariates [17]. Research by Zhao et al. confirmed a significant correlation between elevated TyG index and increased risks of arterial stiffness and nephric microvascular damage among the elderly [21]. Most research exploring the association between IR and vascular damage have been conducted in specific populations, such as elderly, diabetic patients, and specific ethnic group, which has resulted in a lack of evidence from the general population. Yunnan Province, located in southwestern China (Figure S1 A), has a total population of about 46.73 million and is home to 25 ethnic minority groups. The unique geography, climate and diet of Yunnan Province may influence the incidence and prevalence of chronic diseases, vascular damage, and CVDs. Previous research has indicated that the prevalence of hypertension, diabetes, and dyslipidemia in Yunnan Province is higher [28]. However, epidemiological data on IR and vascular damage in this region are limited, with few studies exploring the association between IR and vascular damage. Therefore, this study aimed to investigate the association between macrovascular (PAD) and microvascular damage (CKD and increased albuminuria) and IR (assessed by the TyG index) within a community-dwelling adult population in Yunnan Province. Methods Study sites, participants, and sampling The data for this study were derived from a cross-sectional study conducted in Yunnan Province between January and December 2021, as a part of the Chinese Cardiovascular Diseases and Risk Factor Surveillance Program. Participants were selected using a stratified multi-stage random sampling method: (a) all counties and districts in Yunnan Province were divided into urban and rural groups; (b) four counties or districts were selected from each group using probability proportional to size method; (c) simple random sampling was used to choose two townships/street in each sampled county/district; (d) for each chosen township/street, three communities were selected using simple random sampling; (e) participants then randomly chosen from the selected communities, ensuring representation based on sex and age structure. The locations of the eight surveyed districts/counties are presented in Figure S1 B. Based on the outlined sampling procedure and the prevalence of hypertension, it was estimated that 9600 participants would be required for the survey. A total of 9,600 participants were surveyed between January and December 2021; however, 741 participants did not meet quality control standards. To mitigate the influence of confounding factors associated with CVDs on the study outcomes, 649 participants with existing CVDs or a family history of CVDs were excluded. Among the remaining 8210 participants, those with missing information or abnormal values for key variables were excluded. A total of 3201 participants were finally included in the current study. The detailed participants selection process is shown in Fig. 1. Fig. 1Flowchart of the study participants enrollment Flowchart of the study participants enrollment Data collection Information on sociodemographic (age, sex, ethnicity, and education attainment, etc.), health-related behaviors (smoking statues, alcohol consumption, and diet habits, etc.), chronic and cardiovascular diseases (time of diseases diagnosis, medication information, and family history of diseases, etc.) were collected through a structured questionnaire completed via face-to-face interviews. Participants also underwent a physical examination, which included the measurement of body weight, body height, waist circumference, and four-limb blood pressure. These examinations were conducted by trained professionals using calibrated instruments. Additionally, blood and urine samples were collected from participants in the fasting state to evaluate fasting blood glucose (FBG), lipid profiles (triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), and low-density lipoprotein cholesterol (LDL-C)), urinary creatinine (UCr), serum creatinine (Scr), etc. Variables and definition In this study, all ethnic groups other than Han Chinese were classified as minorities. The education attainment was classified into two levels: below senior high school (including illiteracy, primary school, and junior high school) and senior high school and above (including senior high school, undergraduate, and tertiary education). The questionnaire assessed participants’ smoking status within the past month, with current smoking defined as having smoked in the past month (smoked but not every day, or smoked every day) [29]. Current alcohol intake was defined as the intake of alcohol at least once a month [29]. Body mass index (BMI) was calculated as weight (kg)/height2 (m2). According to Chinese criteria for adult weight, participants were classified as either normal weight or overweight/obese (BMI ≥ 24 kg/m2) [30]. Central obesity was defined as waist circumference ≥ 90 centimeters in males and ≥ 85 centimeters in females [30]. Hypertension was defined according to criteria of a systolic blood pressure (SBP) ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, currently taking antihypertensive medication, or self-reported diagnosis from a physician [31]. Diabetes was indicated by FBG ≥ 7.0 mmol/L (126 mg/dL), current taking antidiabetic medication, or self-reported diagnosis of diabetes by a physician [32]. Dyslipidemia was determined by one or more of the following criteria: TC ≥ 5.20 mmol/L (201 mg/dL), TG ≥ 1.7 mmol/L (150.62 mg/dL), HDL-C < 1 mmol/L (38.66 mg/dL), LDL-C ≥ 3.4 mmol/L (131.44 mg/dL), self-reported history of dyslipidemia, or currently taking lipid-modifying drugs. The TyG index was calculated as ln[TG (mg/dL) × FBG (mg/dL)/2]. The ankle-brachial index (ABI) was calculated by dividing the SBP in the ankle by the SBP in the ipsilateral arm. Macrovascular damage was defined as either leg had an ABI ≤ 0.9. Participants with an ABI ≥ 1.4 were excluded as these values may be falsely elevated due to severe vascular calcification [33]. In this study, microvascular damage included increased albuminuria and chronic kidney disease (CKD). Increased albuminuria was defined by the urinary albumin to creatinine ratio (UACR) > 30 mg/g. As the number of participants with severely increased albuminuria is small (Table S1), this study does not distinguish between moderately (30 mg/g < UACR ≤ 300 mg/g) and severely (UACR > 300 mg/g) increased albuminuria. CKD was identified based on the estimated glomerular filtration rate (eGFR) ≤ 60 mL/min per 1.73 m2. The eGFR was calculated using the Chinese-abbreviated Modification of Diet in Renal Disease (c-aMDRD) formula: 175×Scr− 1.234×age− 0.179 × 0.79 (if female) [34]. Statistical analysis Continuous variables were presented as mean ± standard deviation (normal distribution) or as median (25th percentile (P25), 75th percentile (P75)) (non-normal distribution). Categorical variables were presented as numbers and percentages. Participants were categorized into four groups based on the quantiles of the TyG index. Differences in participant characteristics among the four groups were compared using analysis of variance, the Kruskal-Wallis rank sum test, the chi-squared test, or the Fisher exact test. The relationship between the TyG index and the presence of macrovascular (ABI ≤ 0.9) and microvascular (increased albuminuria and CKD) damage was evaluated through odds ratio (OR)/adjusted odds ratio (aOR) and 95% confidence interval (95% CI), which were calculated using univariate/multivariate logistic regression models. Factors that were significantly associated with macrovascular and microvascular damage in univariate analyses were included in multivariate analyses. Four models were developed: Model 1: crude model; Model 2: adjusted for demographic characteristics, including age, sex, education attainment, and household income; Model 3: adjusted for demographic characteristics and behavioral risk factors (including current smoking and alcohol intake); Model 4: adjusted for demographic characteristics, behavioral risk factors, and metabolic risk factors (overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia). Subgroup and interaction analyses were performed to explore the association between the TyG index and the presence of macro- and microvascular damage across different characteristics. Restricted cubic splines (RCS) regression with three, four, and five knots were used to flexibly model the association between TyG index and the presence of vascular damage [35, 36]. The RCS model with the smallest Bayesian information criterion (BIC), i.e., the model with three knots at the 10th, 50th, and 90th percentiles, was selected (Table S2). The test for multicollinearity (Table S3) indicated that there was no evidence of significant multicollinearity among the covariates (variance inflation factor < 5) [37]. To ensure the robustness of the findings, several sensitivity analyses were conducted: (a) in Model 4, systolic and diastolic blood pressure were used as continuous variables to replace the binary variable “hypertension”; (b) UACR and eGFR were used as response variable for linear regression models, subgroup analyses, interaction analyses, and RCS regression models; (c) the Chinese-modified CKD-EPI (cCKD-EPI) equation was used to calculate eGFR, and the above regression analyses were then conducted based on the these results to explore the impacts of eGFR calculation equations on the outcomes. All statistical analyses were conducted using R software (version 4.4.1, https://www.r-project.org/), and a two-sided p value less than 0.05 was considered statistically significant. Results Characteristics of participants A total of 3201 participants were involved, of which 51.64% were females, with a median age of 44.00 (25th percentile (P25), 75th percentile (P75): 31.00, 57.00) years and a nonnormal distributed TyG index (median: 8.52; P25, P75: 8.08, 8.98; Table 1, Figure S2). Among the participants, 26.30% were minorities, 40.61% had senior high school and above education attainment, 56.64% had a per capita disposable income ≥ 21,000 yuan. The prevalence rates for current smoking, current alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia were 28.36%, 13.73%, 48.39%, 25.49%, 37.86%, 5.47%, and 53.11%, respectively. The prevalence rates for increased albuminuria, chronic kidney disease (CKD), and ABI ≤ 0.9 were 10.65%, 2.53%, and 10.62%, respectively. Higher proportions of older adults, males, low education, low household income, smoking, alcohol consumption, overweight/obesity, central obesity, hypertension, diabetes, dyslipidemia, increased albuminuria, CKD, and an ABI ≤ 0.9 were observed among participants with a high TyG index (p < 0.001). Table 1Characteristics of participants stratified by quartiles of the TyG index*CharacteristicsOverallQ1[6.81, 8.08)Q2[8.08, 8.52)Q3[8.52, 8.98)Q4[8.98, 10.56]p value†N3201800800800801TyG index8.52 (8.08, 8.98)7.81 (7.61, 7.96)8.31 (8.19, 8.42)8.74 (8.63, 8.84)9.31 (9.11, 9.60)< 0.001Age (years)44.00 (31.00, 57.00)33.00 (22.00, 47.00)44.00 (28.75, 57.00)48.50 (36.00, 59.00)49.00 (37.00, 59.00)< 0.001Age group, ≥ 60 years625 (19.53)100 (12.50)158 (19.75)183 (22.88)184 (22.97)< 0.001Sex, female1653 (51.64)478 (59.75)449 (56.12)399 (49.88)327 (40.82)< 0.001Ethnicity, Minorities842 (26.30)225 (28.12)196 (24.50)200 (25.00)221 (27.59)0.251Education attainment, senior high school and above1300 (40.61)464 (58.00)326 (40.75)251 (31.37)259 (32.33)< 0.001Annual per capita household income, ≥ 21,000 yuan1813 (56.64)488 (61.00)440 (55.00)436 (54.50)449 (56.05)0.034Current smoking900 (28.36)177 (22.32)212 (26.73)224 (28.14)287 (36.24)< 0.001Current alcohol intake436 (13.73)87 (10.97)85 (10.72)129 (16.21)135 (17.00)< 0.001BMI (kg/m2)23.81 (21.20, 26.46)21.53 (19.55, 24.40)22.92 (20.71, 25.19)24.65 (22.24, 27.04)25.79 (23.64, 28.20)< 0.001Overweight/Obesity1549 (48.39)222 (27.75)287 (35.88)464 (58.00)576 (71.91)< 0.001Central obesity816 (25.49)85 (10.62)139 (17.38)235 (29.38)357 (44.57)< 0.001SBP (mmHg)125 (114, 142)116 (107, 129)122 (113, 137)131 (118, 146)135 (121, 147)< 0.001DBP (mmHg)80 (72, 88)75 (69, 83)78 (71, 86)82 (75, 89)84 (77, 92)< 0.001Hypertension1212 (37.86)160 (20.00)255 (31.87)374 (46.75)423 (52.81)< 0.001FBG (mg/dL)87.66 (77.04, 99.72)78.84 (71.28, 89.46)85.05 (77.04, 95.22)90.00 (79.74, 101.38)97.92 (85.50, 109.08)< 0.001Diabetes175 (5.47)11 (1.38)23 (2.88)37 (4.62)104 (12.98)< 0.001TG (mg/dL)112.52 (77.08, 174.54)60.25 (49.62, 70.88)93.92 (82.40, 106.54)139.10 (119.39, 158.82)235.68 (194.03, 303.01)< 0.001TC (mg/dL)182.48 (158.51, 208.76)161.99 (141.88, 184.41)177.45 (155.41, 200.65)192.14 (167.69, 212.63)202.19 (176.68, 228.48)< 0.001HDL-C (mg/dL)55.67 (47.55, 64.56)61.08 (52.96, 70.75)58.38 (51.03, 65.72)53.74 (47.17, 61.47)48.71 (41.75, 56.44)< 0.001LDL-C (mg/dL)110.57 (90.46, 133.76)93.17 (77.32, 111.34)107.86 (89.30, 128.45)120.62 (101.19, 140.82)124.10 (102.06, 146.13)< 0.001Dyslipidemia1700 (53.11)128 (16.00)249 (31.13)531 (66.38)792 (98.88)< 0.001UACR (mg/g)6.00 (3.37, 12.96)5.25 (3.13, 10.37)5.88 (3.29, 12.06)6.18 (3.38, 13.01)6.79 (3.72, 18.39)< 0.001Increased albuminuria341 (10.65)57 (7.12)81 (10.12)84 (10.50)119 (14.86)< 0.001eGFR (mL/min per 1.73 m2)98.62 (84.59, 121.22)102.18 (89.24, 125.04)99.43 (84.28, 123.17)97.00 (83.42, 119.49)94.83 (81.53, 118.45)< 0.001CKD81 (2.53)3 (0.38)17 (2.12)25 (3.12)36 (4.49)< 0.001ABI ≤ 0.9340 (10.62)49 (6.12)83 (10.38)99 (12.38)109 (13.61)< 0.001* Numbers were given as n (%) or median (P25, P75)† Differences among the four groups were compared using the Kruskal-Wallis rank sum test or the chi-squared test Table 1Characteristics of participants stratified by quartiles of the TyG index*CharacteristicsOverallQ1[6.81, 8.08)Q2[8.08, 8.52)Q3[8.52, 8.98)Q4[8.98, 10.56]p value†N3201800800800801TyG index8.52 (8.08, 8.98)7.81 (7.61, 7.96)8.31 (8.19, 8.42)8.74 (8.63, 8.84)9.31 (9.11, 9.60)< 0.001Age (years)44.00 (31.00, 57.00)33.00 (22.00, 47.00)44.00 (28.75, 57.00)48.50 (36.00, 59.00)49.00 (37.00, 59.00)< 0.001Age group, ≥ 60 years625 (19.53)100 (12.50)158 (19.75)183 (22.88)184 (22.97)< 0.001Sex, female1653 (51.64)478 (59.75)449 (56.12)399 (49.88)327 (40.82)< 0.001Ethnicity, Minorities842 (26.30)225 (28.12)196 (24.50)200 (25.00)221 (27.59)0.251Education attainment, senior high school and above1300 (40.61)464 (58.00)326 (40.75)251 (31.37)259 (32.33)< 0.001Annual per capita household income, ≥ 21,000 yuan1813 (56.64)488 (61.00)440 (55.00)436 (54.50)449 (56.05)0.034Current smoking900 (28.36)177 (22.32)212 (26.73)224 (28.14)287 (36.24)< 0.001Current alcohol intake436 (13.73)87 (10.97)85 (10.72)129 (16.21)135 (17.00)< 0.001BMI (kg/m2)23.81 (21.20, 26.46)21.53 (19.55, 24.40)22.92 (20.71, 25.19)24.65 (22.24, 27.04)25.79 (23.64, 28.20)< 0.001Overweight/Obesity1549 (48.39)222 (27.75)287 (35.88)464 (58.00)576 (71.91)< 0.001Central obesity816 (25.49)85 (10.62)139 (17.38)235 (29.38)357 (44.57)< 0.001SBP (mmHg)125 (114, 142)116 (107, 129)122 (113, 137)131 (118, 146)135 (121, 147)< 0.001DBP (mmHg)80 (72, 88)75 (69, 83)78 (71, 86)82 (75, 89)84 (77, 92)< 0.001Hypertension1212 (37.86)160 (20.00)255 (31.87)374 (46.75)423 (52.81)< 0.001FBG (mg/dL)87.66 (77.04, 99.72)78.84 (71.28, 89.46)85.05 (77.04, 95.22)90.00 (79.74, 101.38)97.92 (85.50, 109.08)< 0.001Diabetes175 (5.47)11 (1.38)23 (2.88)37 (4.62)104 (12.98)< 0.001TG (mg/dL)112.52 (77.08, 174.54)60.25 (49.62, 70.88)93.92 (82.40, 106.54)139.10 (119.39, 158.82)235.68 (194.03, 303.01)< 0.001TC (mg/dL)182.48 (158.51, 208.76)161.99 (141.88, 184.41)177.45 (155.41, 200.65)192.14 (167.69, 212.63)202.19 (176.68, 228.48)< 0.001HDL-C (mg/dL)55.67 (47.55, 64.56)61.08 (52.96, 70.75)58.38 (51.03, 65.72)53.74 (47.17, 61.47)48.71 (41.75, 56.44)< 0.001LDL-C (mg/dL)110.57 (90.46, 133.76)93.17 (77.32, 111.34)107.86 (89.30, 128.45)120.62 (101.19, 140.82)124.10 (102.06, 146.13)< 0.001Dyslipidemia1700 (53.11)128 (16.00)249 (31.13)531 (66.38)792 (98.88)< 0.001UACR (mg/g)6.00 (3.37, 12.96)5.25 (3.13, 10.37)5.88 (3.29, 12.06)6.18 (3.38, 13.01)6.79 (3.72, 18.39)< 0.001Increased albuminuria341 (10.65)57 (7.12)81 (10.12)84 (10.50)119 (14.86)< 0.001eGFR (mL/min per 1.73 m2)98.62 (84.59, 121.22)102.18 (89.24, 125.04)99.43 (84.28, 123.17)97.00 (83.42, 119.49)94.83 (81.53, 118.45)< 0.001CKD81 (2.53)3 (0.38)17 (2.12)25 (3.12)36 (4.49)< 0.001ABI ≤ 0.9340 (10.62)49 (6.12)83 (10.38)99 (12.38)109 (13.61)< 0.001* Numbers were given as n (%) or median (P25, P75)† Differences among the four groups were compared using the Kruskal-Wallis rank sum test or the chi-squared test Characteristics of participants stratified by quartiles of the TyG index* * Numbers were given as n (%) or median (P25, P75) † Differences among the four groups were compared using the Kruskal-Wallis rank sum test or the chi-squared test Factors influencing the presence of macro- and microvascular damage were analyzed by univariate logistic regression The results of the univariate logistic regression models indicated significant associations between older adults, low educational attainment, overweight/obesity, hypertension, dyslipidemia, and higher TyG index with the presence of ABI ≤ 0.9, increased albuminuria, and CKD (p < 0.05, Table S4). Participants who were male, had low household income, current alcohol intake (OR: 2.03, 95% CI: 1.53 to 2.66), central obesity (OR: 1.33, 95% CI: 1.04 to 1.70), and diabetes (OR: 2.06, 95% CI: 1.36 to 3.02) demonstrated significant associations with the presence of ABI ≤ 0.9. Participants who currently drank alcohol (OR: 1.37, 95% CI: 1.01 to 1.85), were central obesity (OR: 1.59, 95% CI: 1.25 to 2.02), and had diabetes (OR: 2.40, 95% CI: 1.62 to 3.48) had significant associations with the presence of increased albuminuria. Participants with low household income were related to the presence of CKD (p = 0.026). Relationship between the TyG index and the presence of macro- and microvascular damage Four logistic regression models were developed to explore the relationships between the TyG index and the presence of ABI ≤ 0.9, increased albuminuria, and CKD (Table 2). In the univariate logistic regression model (Model 1), a 1-unit increment in the TyG index was associated with a 60.75% increased odds of ABI ≤ 0.9 (OR: 1.61, 95% CI: 1.35 to 1.91), a 59.35% increased odds of increased albuminuria (OR: 1.59, 95% CI: 1.34 to 1.89), and a 129.60% increased odds of CKD (OR: 2.30, 95% CI: 1.65 to 3.20). After adjustment for demographic characteristics (Model 2) and behavioral risk factors (Model 3), each unit increase in TyG index was associated with a 1.37- and 1.39-fold increased odds of ABI ≤ 0.9, a 1.57- and 1.60-fold increased odds of increased albuminuria, and a 2.22- and 2.25-fold increased odds of CKD (p < 0.05). In Model 4, which adjusted for demographic, behavioral, and metabolic actors, the TyG index was not associated with the presence of ABI ≤ 0.9, but each unit increase in the TyG index was associated with a 1.37-fold higher odds of increased albuminuria and a 2.03-fold higher odds of CKD. Additionally, we converted the TyG index from a continuous variable into a categorical variable based on quartiles (Table 2). In crude model (Model 1), Model 2, and Model 3, compared to the lowest quartile (Q1), the odds of ABI ≤ 0.9, increased albuminuria, and CKD significantly increased for the subsequent quartiles (Q2, Q3, and Q4) (p < 0.05, and p for trend < 0.05). After adjusting for demographic characteristics, behavioral and metabolic factors (Model 4), participants in the high TyG index quartile had a higher odds of ABI ≤ 0.9, but the difference was not statistically significant (p > 0.05, p for trend = 0.162); those in Q4 experienced a 1.68-fold higher odds of increased albuminuria (p = 0.022, p for trend = 0.041); those in the high TyG index quantile had a significantly higher odds of CKD (p < 0.05, p for trend < 0.001). To further explore the relationship between the TyG index and the presence of ABI ≤ 0.9, increased albuminuria, and CKD, subgroup analyses and interaction analyses were conducted across different demographic characteristics, behavioral and metabolic factors (Fig. 2). The relationship between the TyG index and the presence of ABI ≤ 0.9 and increased albuminuria were not affected by demographic characteristics, behavioral and metabolic factors (p for interaction > 0.05, Fig. 2). In contrast, the association between the TyG index and the presence of CKD was varied among participants with different statuses of overweight/obesity (p for interaction = 0.009) and dyslipidemia (p for interaction = 0.008). For each unit increase in the TyG index, the odds of CKD manifestation was higher in normal-weight participants (OR: 5.36, 95% CI: 2.38 to 12.66) than in overweight/obesity participants (OR: 1.23, 95% CI: 0.69 to 2.20), and the odds of CKD manifestation was higher in participants with normal lipids (OR: 8.60, 95% CI: 2.32 to 37.61) than in those with dyslipidemia (OR: 1.60, 95% CI: 0.97 to 2.65). Table 2Relationships between the TyG index and the presence of macrovascular (ABI ≤ 0.9) and microvascular (increased albuminuria and CKD) damage in different modelsModels, characteristicsABI ≤ 0.9Increased albuminuriaCKDOR/aOR (95% CI)p valueOR/aOR (95% CI)p valueOR/aOR (95% CI)p valueModel 1 TyG index (per unit)1.61 (1.35 to 1.91)< 0.0011.59 (1.34 to 1.89)< 0.0012.30 (1.65 to 3.20)< 0.001 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.77 (1.23 to 2.58)0.0021.47 (1.03 to 2.10)0.0335.77 (1.93 to 24.78)0.005 Q32.16 (1.52 to 3.12)< 0.0011.53 (1.08 to 2.18)0.0188.57 (2.99 to 36.10)< 0.001 Q42.41 (1.71 to 3.46)< 0.0012.27 (1.64 to 3.19)< 0.00112.50 (4.49 to 51.99)< 0.001 p for trend< 0.001< 0.001< 0.001Model 2 TyG index (per unit)1.37 (1.14 to 1.65)< 0.0011.57 (1.31 to 1.89)< 0.0012.22 (1.56 to 3.17)< 0.001 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.47 (1.01 to 2.15)0.0441.36 (0.96 to 1.96)0.0905.02 (1.66 to 21.64)0.011 Q31.63 (1.13 to 2.36)0.0101.39 (0.97 to 2.00)0.0757.14 (2.46 to 30.29)0.001 Q41.78 (1.25 to 2.58)0.0022.13 (1.52 to 3.01)< 0.00110.49 (3.72 to 43.94)< 0.001 p for trend0.002< 0.001< 0.001Model 3 TyG index (per unit)1.39 (1.16 to 1.68)< 0.0011.60 (1.33 to 1.92)< 0.0012.25 (1.57 to 3.21)< 0.001 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.55 (1.07 to 2.28)0.0231.35 (0.94 to 1.95)0.1025.05 (1.67 to 21.79)0.010 Q31.67 (1.16 to 2.44)0.0071.38 (0.96 to 1.99)0.0817.12 (2.46 to 30.20)0.001 Q41.88 (1.31 to 2.73)< 0.0012.18 (1.55 to 3.09)< 0.00110.62 (3.76 to 44.48)< 0.001 p for trend0.001< 0.001< 0.001Model 4 TyG index (per unit)1.21 (0.95 to 1.54)0.1271.37 (1.07 to 1.74)0.0122.03 (1.29 to 3.24)0.003 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.45 (0.99 to 2.14)0.0611.26 (0.87 to 1.83)0.2224.88 (1.61 to 21.14)0.012 Q31.40 (0.94 to 2.11)0.1031.14 (0.77 to 1.71)0.5076.48 (2.13 to 28.24)0.003 Q41.47 (0.94 to 2.31)0.0971.68 (1.08 to 2.62)0.0229.47 (2.95 to 42.89)< 0.001 p for trend0.1620.041< 0.001Model 1: we did not adjust other covariates. Model 2: we adjusted age, sex, education attainment, and annual per capita household income. Model 3: we adjusted age, sex, education attainment, annual per capita household income, and current smoking and alcohol intake. Model 4: we adjusted age, sex, education attainment, annual per capita household income, current smoking and alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia. OR=Odds ratio, aOR=Adjusted odds ratio, 95% CI: 95% confidence interval Table 2Relationships between the TyG index and the presence of macrovascular (ABI ≤ 0.9) and microvascular (increased albuminuria and CKD) damage in different modelsModels, characteristicsABI ≤ 0.9Increased albuminuriaCKDOR/aOR (95% CI)p valueOR/aOR (95% CI)p valueOR/aOR (95% CI)p valueModel 1 TyG index (per unit)1.61 (1.35 to 1.91)< 0.0011.59 (1.34 to 1.89)< 0.0012.30 (1.65 to 3.20)< 0.001 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.77 (1.23 to 2.58)0.0021.47 (1.03 to 2.10)0.0335.77 (1.93 to 24.78)0.005 Q32.16 (1.52 to 3.12)< 0.0011.53 (1.08 to 2.18)0.0188.57 (2.99 to 36.10)< 0.001 Q42.41 (1.71 to 3.46)< 0.0012.27 (1.64 to 3.19)< 0.00112.50 (4.49 to 51.99)< 0.001 p for trend< 0.001< 0.001< 0.001Model 2 TyG index (per unit)1.37 (1.14 to 1.65)< 0.0011.57 (1.31 to 1.89)< 0.0012.22 (1.56 to 3.17)< 0.001 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.47 (1.01 to 2.15)0.0441.36 (0.96 to 1.96)0.0905.02 (1.66 to 21.64)0.011 Q31.63 (1.13 to 2.36)0.0101.39 (0.97 to 2.00)0.0757.14 (2.46 to 30.29)0.001 Q41.78 (1.25 to 2.58)0.0022.13 (1.52 to 3.01)< 0.00110.49 (3.72 to 43.94)< 0.001 p for trend0.002< 0.001< 0.001Model 3 TyG index (per unit)1.39 (1.16 to 1.68)< 0.0011.60 (1.33 to 1.92)< 0.0012.25 (1.57 to 3.21)< 0.001 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.55 (1.07 to 2.28)0.0231.35 (0.94 to 1.95)0.1025.05 (1.67 to 21.79)0.010 Q31.67 (1.16 to 2.44)0.0071.38 (0.96 to 1.99)0.0817.12 (2.46 to 30.20)0.001 Q41.88 (1.31 to 2.73)< 0.0012.18 (1.55 to 3.09)< 0.00110.62 (3.76 to 44.48)< 0.001 p for trend0.001< 0.001< 0.001Model 4 TyG index (per unit)1.21 (0.95 to 1.54)0.1271.37 (1.07 to 1.74)0.0122.03 (1.29 to 3.24)0.003 TyG index quartile Q11.00 (reference)-1.00 (reference)-1.00 (reference)- Q21.45 (0.99 to 2.14)0.0611.26 (0.87 to 1.83)0.2224.88 (1.61 to 21.14)0.012 Q31.40 (0.94 to 2.11)0.1031.14 (0.77 to 1.71)0.5076.48 (2.13 to 28.24)0.003 Q41.47 (0.94 to 2.31)0.0971.68 (1.08 to 2.62)0.0229.47 (2.95 to 42.89)< 0.001 p for trend0.1620.041< 0.001Model 1: we did not adjust other covariates. Model 2: we adjusted age, sex, education attainment, and annual per capita household income. Model 3: we adjusted age, sex, education attainment, annual per capita household income, and current smoking and alcohol intake. Model 4: we adjusted age, sex, education attainment, annual per capita household income, current smoking and alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia. OR=Odds ratio, aOR=Adjusted odds ratio, 95% CI: 95% confidence interval Relationships between the TyG index and the presence of macrovascular (ABI ≤ 0.9) and microvascular (increased albuminuria and CKD) damage in different models Model 1: we did not adjust other covariates. Model 2: we adjusted age, sex, education attainment, and annual per capita household income. Model 3: we adjusted age, sex, education attainment, annual per capita household income, and current smoking and alcohol intake. Model 4: we adjusted age, sex, education attainment, annual per capita household income, current smoking and alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia. OR=Odds ratio, aOR=Adjusted odds ratio, 95% CI: 95% confidence interval Fig. 2Subgroup analyses of the relationships between the TyG index and the presence of macrovascular (ABI ≤ 0.9) and microvascular (increased albuminuria and CKD) damage. The black dots were adjusted odds ratios (aORs) for each unit change in the TyG index, and the error bars were 95% confidence intervals (95% CIs). Gray dashed vertical lines indicated an aOR of 1.00 (reference). These analyses adjusted for age, sex, education attainment, annual per capita household income, current smoking and alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia Subgroup analyses of the relationships between the TyG index and the presence of macrovascular (ABI ≤ 0.9) and microvascular (increased albuminuria and CKD) damage. The black dots were adjusted odds ratios (aORs) for each unit change in the TyG index, and the error bars were 95% confidence intervals (95% CIs). Gray dashed vertical lines indicated an aOR of 1.00 (reference). These analyses adjusted for age, sex, education attainment, annual per capita household income, current smoking and alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia Nonlinear relationship between the TyG index and the presence of macro- and microvascular damage Nonlinear association were observed between the TyG index and the presence of ABI ≤ 0.9 (p = 0.037, p for nonlinear = 0.033) and CKD (p = 0.002, p for nonlinear = 0.010) (Fig. 3). Specifically, the odds of ABI ≤ 0.9 and CKD increased nonlinearly with the advancing TyG index when it was below 8.91, but remained stable when the TyG index exceeded 8.91. The likelihood of increased albuminuria rose with an advancing TyG index (p = 0.026), but this relationship was not nonlinear (p for nonlinear = 0.352). Fig. 3The restricted cubic spline analysis examined the association between the TyG index and the presence of ABI ≤ 0.9 (A), increased albuminuria (B), and CKD (C). The reference values (gray dashed vertical line) were set at a TyG index of 8.91, and the knots were set at 10th, 50th, and 90th percentiles of the TyG index. The models were adjusted for age, sex, education attainment, annual per capita household income, current smoking and alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia. Grey dashed horizontal lines indicated an aOR of 1.00 (reference), black solid lines represented adjusted odds ratios, and shadow areas represented the 95% confidence intervals The restricted cubic spline analysis examined the association between the TyG index and the presence of ABI ≤ 0.9 (A), increased albuminuria (B), and CKD (C). The reference values (gray dashed vertical line) were set at a TyG index of 8.91, and the knots were set at 10th, 50th, and 90th percentiles of the TyG index. The models were adjusted for age, sex, education attainment, annual per capita household income, current smoking and alcohol intake, overweight/obesity, central obesity, hypertension, diabetes, and dyslipidemia. Grey dashed horizontal lines indicated an aOR of 1.00 (reference), black solid lines represented adjusted odds ratios, and shadow areas represented the 95% confidence intervals Sensitivity analysis Firstly, adjusting the multivariate model by replacing the presence of hypertension with systolic blood pressure (Model S1, Table S5) and diastolic blood pressure (Model S2, Table S5) yielded results consistent with Model 4 in the main analysis (Table 2). The TyG index was not significantly associated with ABI < 0.9, while an elevated TyG index was significantly associated with higher odds of increased albuminuria and CKD (aOR > 1, p < 0.05). Secondly, we employed UACR and eGFR as response variables for linear regression models, subgroup and interaction analyses, and restricted cubic spline regression models. In the univariate linear regression model (Model 1, Table S6), as well as in models adjusted for demographic or/and behavioral factors (Model 2 and 3, Table S6), an elevated TyG index was positively correlated with increased UACR and decreased eGFR. However, in the model adjusted for demographic characteristics, behavioral, and metabolic factors (Model 4, Table S6), the elevated TyG index was only positively associated with increased UACR. Subgroup and interaction analyses suggested that the association between the TyG index and UACR was influenced by age (p for interaction < 0.001) and diabetes (p for interaction < 0.001); the association between the TyG index and eGFR was affected by education level (p for interaction = 0.048), alcohol consumption (p for interaction = 0.006), and diabetes (p for interaction = 0.009, Table S7). There was no evidence of a nonlinear association between the TyG index and either UACR (p for nonlinear = 0.253) or eGFR (p for nonlinear = 0.567, Figure S3). Lastly, we evaluated the impact of eGFR estimates derived from different equations on the results. The eGFR calculated using the cCKD-EPI equation was lower than that obtained from the c-aMDRD equation, resulting in a CKD (cCKD-EPI) prevalence rate of 4.75% (Table S8). The findings from multivariate regression (Table S9), subgroup and interaction analysis (Table S10), and RCS regression (Figure S4) regarding the association between the TyG index and CKD (cCKD-EPI) were consistent with those of the main analysis (c-aMDRD). Discussion This cross-sectional study systematically revealed a complex association between IR (measured by the TyG index) and vascular damage within a community-dwelling adult population in Yunnan, China. An increased TyG index was significantly associated with a higher odds of peripheral arterial disease (ABI ≤ 0.9), CKD, and increased albuminuria, and these associations remained significant after adjusting for demographic characteristics, behavioral and metabolic factors. The relationship between CKD and IR was affected by overweight/obesity and dyslipidemia. Additionally, the odds of ABI ≤ 0.9 and CKD increased nonlinearly with the increasing TyG index when it was below 8.91, but remained constant when the TyG index exceeded this threshold. IR is recognized as an important risk factor for CVDs [17–19], PAD [23], CKD [25], and other diseases [21, 22, 26]. Because of the complexity and expense associated with directly measuring IR, several surrogate markers have been proposed, one of which is the TyG index. This index has demonstrated a strong correlation with the HEC test [15, 16]. Therefore, this study used the TyG index as a biomarker for IR. Several studies have established the association between the TyG index and macrovascular damage, including arterial stiffness [21, 22], PAD [23], and carotid atherosclerosis [24]. Liu et al. demonstrated in a general population cohort that an elevated TyG index was associated with an increased likelihood of CVD among the U.S. population aged ≤ 65 years with prediabetes and diabetes [38]. Notably, our study extends these findings by simultaneously evaluating both macrovascular (ABI ≤ 0.9) and microvascular (increased albuminuria, CKD) endpoints in a community-based sample. Consistent with findings from a prospective cohort study [23], this study found that the TyG index was independently associated with PAD (ABI ≤ 0.9), supporting its value for the early detection of PAD. Furthermore, CKD and increased albuminuria were linked to a higher risk of CVD, yet few studies examined the association between IR and nephric microvascular damage. The results of this study indicated that the TyG index was an independent risk factor for CKD and increased albuminuria, suggesting that IR was independently associated with an increased odds of nephric microvascular damage. In addition, this study showed that the odds of the presence of PAD and CKD increased nonlinearly as the TyG index increased, particularly when the TyG index was below 8.91. This relationship establishes a reference threshold for the early detection of vascular damage using the TyG index. Future studies are required to further investigate the role of the TyG index in the diagnosis of vascular damage. In both this study and previous studies, subgroup analyses and interaction analyses were used to investigate the robustness of the association between IR and vascular damage. The findings from Gao et al.’s prospective cohort study indicated that the relationship between the TyG index and PAD remained consistent after stratifying participants by age, sex, race, smoking status, BMI, hypertension, and diabetes [23]. Furthermore, Hsu et al.’s prospective cohort analysis involving 738 individuals with type 2 diabetes confirmed that the association between the TyG index and microalbuminuria was unaffected by metabolic status [26]. Consistent with these studies, this study also demonstrated that the association between the TyG index and an ABI ≤ 0.9, as well as increased albuminuria, was unaffected by demographic characteristics, behavioral and metabolic factors. The association between the TyG index and CKD appears to vary depending on metabolic status. Research by Chen et al. revealed that the relationship between the TyG index and CKD was influenced by the presence of hypertension [25]. This study found that the association between IR and CKD was influenced by dyslipidemia and overweight/obesity. These results highlight the need for targeted intervention measures to mitigate CKD risks among populations affected by hypertension, diabetes, and obesity. Notably, this study found that IR significantly increased the odds of CKD in individuals with normal lipid levels. This finding may be related to the complex relationship between lipid metabolism and vascular damage, with previous studies suggesting that normal/high levels of HDL may promote endothelial progenitor cells senescence and reduce tube formation, which in turn increases vascular damage [39]. In addition, patients with dyslipidemia in this cross-sectional study may maintain normal lipid levels by the use of lipid-modifying drugs, potentially leading to the reversed result. Therefore, prospective studies are necessary to further investigate the causal relationships among IR, dyslipidemia, and vascular damage. Strengths and limitations This study used a thorough analytical approach involving large samples, various indicators and different statistical methods to clearly demonstrate the association between the TyG index and vascular damage, which provided scientific evidence for its application in clinical practice and public health policy. Findings of the study suggested that the TyG index, as a cost-effective biomarker, could be integrated into routine clinical assessments and preventive care strategies. This would enable the identification of individuals at risk for vascular damage and facilitate the stratification of patients based on their TyG index. Consequently, it would allow targeted interventions and therapeutic measures to be adopted, focusing on aspects such as diet, exercise, and medication therapy. The following limitations should be considered when interpreting the results of this study. First, this was a cross-sectional study, which limited the ability to establish a causal relationship between the TyG index and vascular damage. In addition, the effect of dyslipidemia on the association between the TyG index and CKD may have been reversed as a result of the initiation of medication and behavioral lifestyle changes in the dyslipidemic patients after diagnosis. Second, this study was only conducted among residents of Yunnan Province, so extrapolating these results to other populations and regions should be done with caution. Last, inflammation-induced endothelial dysfunction was also a possible underlying mechanism connecting IR to vascular damage. However, because inflammation-related markers (such as high-sensitivity C-reactive protein) and complete blood count (neutrophils, lymphocyte, monocyte, platelet, etc.) were not included in the survey, this study could not examine the effects of inflammation on the association between the TyG index and vascular damage. There is a significant need for further studies to explore the relationship between inflammation and the TyG index and their impacts on vascular damage. Conclusions An elevated TyG index was independently associated with increased odds of both macro- and microvascular damage, and these associations were nonlinear, suggesting that the TyG index may potentially serve as a biomarker for vascular damage. The association between insulin resistance and CKD was influenced by overweight/obesity and dyslipidemia, suggesting the importance of early management and intervention in these populations to reduce the disease burden associated with CKD. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1