Skip to main content
Download PDF
Download ePub

Trajectories of triglyceride-glucose index changes and their association with all-cause and cardiovascular mortality: a competing risk analysis

Cardiovascular Diabetology volume 23, Article number: 364 (2024) Cite this article

Abstract

Background

The association between changes in insulin resistance, reflected by the triglyceride-glucose (TyG) index, and mortality remains unclear. This study investigated whether longitudinal trajectories of TyG index changes are associated with all-cause and cardiovascular disease (CVD) mortality.

Methods

This retrospective cohort study analyzed data from 233,546 adults aged ≥ 19 years from the Korea National Health Insurance Service-National Sample Cohort. Participants were categorized as having increasing, stable, or decreasing TyG index changes during a 4-year exposure period (2009–2014). Mortality outcomes were assessed during an 8.13-year follow-up period (2015–2021). Cox proportional hazards regression and competing risk analysis were used to evaluate all-cause and CVD mortality.

Results

A total of 7918 mortality events, including 651 CVD deaths, were recorded. Compared with the stable group, adjusted hazard ratios for all-cause mortality were 1.09 (95% CI 1.03–1.15) in the increasing group and 1.23 (95% CI 1.01–1.50) for CVD mortality. An increased TyG index was significantly associated with all-cause mortality in individuals aged < 50 years; men; and individuals with obesity, hypertension, diabetes, and/or dyslipidemia. For CVD mortality, significant associations were found in individuals aged 50–69 years, with obesity, with diabetes, or without dyslipidemia.

Conclusion

An increasing TyG index from baseline during follow-up was independently associated with higher risks of all-cause and CVD mortality. Serial monitoring of TyG index changes could enhance risk stratification and inform targeted interventions to reduce insulin resistance, and ultimately lower mortality risk.

Introduction

Cardiovascular disease (CVD) remains the leading cause of mortality globally, accounting for 34.9% of all deaths in 2022 [1]. In 2021, ischemic heart disease had an age-standardized mortality rate of 108.7, making it the leading cause of death globally, whereas stroke ranked third with a mortality rate of 87.4 [2]. Insulin resistance (IR) is a crucial driver of atherosclerosis and cardiovascular disease (CVD) progression [3, 4]. The hyperinsulinemic–euglycemic clamp is the gold standard for measuring IR, but its complexity limits its clinical use. The triglyceride-glucose (TyG) index has recently gained attention as a simpler, noninvasive surrogate marker for IR [5]. Studies suggest that the TyG index may outperform homeostatic model assessment for IR (HOMA-IR) in predicting cardiovascular and metabolic outcomes because it reflects lipid and glucose metabolism, thereby providing a more comprehensive view of metabolic health [6,7,8,9].

To date, most studies have relied on single-point measurements of IR, which fail to capture dynamic changes over time. Similar to how blood pressure changes occur over a lifespan [10, 11], IR can also increase, remain stable, or decrease over time. Therefore, tracking changing trends of IR could reflect CVD risk, thereby improving risk stratification and potentially more targeted interventions. Some studies indicate that increasing HOMA-IR trajectories are correlated with higher CVD risk and mortality [12], and other studies have explored TyG index changes by using two timepoints [13]. However, these approaches may overlook significant fluctuations in IR over time. The relationship between long-term TyG index trajectories and CVD outcomes, particularly the impact of decreasing values, remains underexplored.

In this context, we hypothesized that long-term trajectories of TyG index changes are associated with CVD outcomes in the Korean population. By using a large Korean cohort, this study conducted a comprehensive longitudinal analysis to explore how dynamic shifts in TyG index trajectories (i.e., increase, decrease, stable) influence long-term CVD and mortality outcomes.

Methods

Study design and population

This retrospective cohort study utilized data from the Korea National Health Insurance Service-National Sample Cohort (NHIS-NSC) 2.2 database. The Korean National Health Screening program conducts biennial health screenings. In 2009, the program began measuring serum triglyceride levels, which is essential for calculating the TyG index. Baseline data from the 2009–2010 NHIS-NSC comprised information on 363,270 participants, with follow-up data available until December 31, 2021. To identify TyG index trajectories, we required data from at least three timepoints while maximizing the event accrual period. Therefore, we defined 2009–2014 as the exposure period, during which TyG index changing patterns were assessed, and 2015–2021 as the event accrual period, during which mortality outcomes were tracked. Mortality status was determined by linking personal identification key codes generated by the NHIS-NSC with national data sources, including death records from the Korea National Statistical Office.

Figure 1 illustrates the flowchart of the study population. Among the 363,270 participants in the 2009–2010 NHIS-NSC, we selected 234,446 individuals who consecutively participated in the NHIS-NSC during the 2009–2010, 2011–2012, and 2013–2014 periods. We excluded (1) individuals aged < 19 years (n = 196), (2) individuals with missing serum triglyceride data (n = 64), (3) individuals with missing glucose data (n = 6), and (4) individuals who died before 2014 (n = 634). Finally, 233,546 participants were included in the analysis.

Fig. 1
figure 1

Flowchart of the study population. Abbreviations: TyG, triglyceride-glucose; CVD, cardiovascular disease; NHIS-NSC, National health insurance service-national sample cohort

Full size image

This study was conducted in accordance with the STROBE Statement. The Institutional Review Board of Eulji University Hospital (Seoul, Republic of Korea) approved the study (approval no. 2023-12-018). The requirement for informed consent was waived because we used anonymized data provided by the NHIS-NSC database, based on the Personal Data Protection Act guidelines.

Trajectories of TyG index changes

The TyG index was calculated, using the following formula [5]: TyG index = ln [triglyceride (mg/dL) × glucose (mg/dL) ÷ 2]. We classified participants into the increasing (n = 64,415), stable (n = 112,493), and decreasing (n = 56,638) TyG index trajectory groups from baseline through follow-up by using Gaussian finite mixture modeling (Supplementary Table 1 and Fig. 2). Detailed information on the method for trajectory modeling is described in the statistical analysis section.

Fig. 2
figure 2

Changes in the TyG index of each individual during the exposure period. Abbreviations: TyG, triglyceride-glucose

Full size image

Outcomes

The primary outcome was all-cause mortality. Mortality status information was linked to the Korea National Statistical Office, which recorded death codes, based on the International Classification of Diseases, Tenth Revision (ICD-10). The secondary outcome was CVD mortality, which included ischemic heart disease (ICD-10 codes I20–I25) and cerebrovascular disease (ICD-10 codes I60–I69). All mortality events between January 2015 and December 2021 were recorded.

Covariates

Height and body weight were measured to the nearest 0.001 m and 0.1 kg, respectively. The body mass index (BMI) was calculated by dividing body weight by height squared (kg/m2), with obesity defined as a BMI ≥ 25 kg/m2 [14]. After at least 5 min of resting, the systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured in the seated position. The levels of fasting blood glucose (FBG), serum creatinine, total cholesterol, triglyceride, and high-density lipoprotein cholesterol (HDL-C) were measured. The estimated glomerular filtration rate (eGFR) was calculated using the Modification of Diet in Renal Disease Study equation [15]. Participants were classified as current smokers or nonsmokers, based on their smoking status, and as current drinkers or nondrinkers, based on their alcohol consumption status. Regular exercise was defined as moderate physical activity ≥ 150 min/week or vigorous physical activity ≥ 75 min/week. A diagnosis of diabetes mellitus (DM) was based on the presence of at least one of the following criteria: (1) FBG ≥ 126 mg/dL, (2) current treatment with DM medication, or (3) ICD-10 codes E11–E14. Participants were considered to have hypertension (HTN) if they met at least one of the following criteria: (1) SBP ≥ 140 mmHg, (2) DBP ≥ 90 mmHg, (3) current treatment with HTN medication, or (4) ICD-10 codes I10–I13. Dyslipidemia (DLD) was defined as the presence of at least one of the following criteria: (1) serum total cholesterol ≥ 240 mg/dL, (2) current treatment with DLD medication, or (3) ICD-10 code E78.

Statistical analysis

Gaussian finite mixture modeling was performed to identify different trajectories of changes in the TyG index from baseline TyG index values during the median 4.0-year exposure period by using the command ‘lcMethodMclust’ in R package ‘latrend’ [16]. This method employs the expectation–maximization (EM) algorithm to estimate the parameters of the mixture model iteratively. The EM algorithm alternates between assigning individuals to latent trajectory groups (i.e., expectation step) and updating group-specific parameters to maximize the likelihood of the observed data (i.e., maximization step). This approach allows identifying underlying patterns and heterogeneity in the longitudinal changes of the TyG index, thereby facilitating a clearer understanding of its variations across different groups within the study population. The criteria for selecting the optimal number of groups were (1) a low Bayesian information criterion, (2) minimum group size exceeding 20% of the total sample, and (3) at least one group with an average posterior probability assignment of 90% or higher.

All data are presented as the mean ± standard deviation or as the median (interquartile range [IQR]: 25th and 75th percentiles) for continuous variables and as the number (percentage [%]) for categorical variables. For group comparisons, analysis of variance was used for continuous variables that satisfied the normality assumption, whereas the Kruskal–Wallis test was used for variables that did not. The chi-square test was applied to categorical variables. An adjusted survival curve was drawn, depicting the survival probability, adjusted for the TyG index at baseline. Additionally, we created an adjusted survival curve by using a proportional subdistribution hazard model to illustrate the cumulative incidence rates of CVD mortality, based on the TyG index trajectory groups, which were also adjusted for the TyG index at baseline [17]. For the primary outcome, Cox proportional hazards regression analysis was conducted to estimate the hazard ratio (HR) with a 95% confidence interval (CI) for the all-cause mortality of the increasing and decreasing TyG index trajectory groups, compared with that of the stable group. The Cox proportional hazards assumption using the Schoenfeld residuals test confirmed that the assumption was met (P = 0.71).

For the secondary outcome, competing risk analysis was conducted by using the Fine and Gray model to estimate the subdistribution HR with a 95% CI for CVD death, setting non-CVD mortality as a competing risk. The TyG index at baseline was adjusted in Model 1. Age, sex, BMI, and the TyG index at baseline were adjusted in Model 2. In Model 3, smoking status, drinking status, and regular exercise were adjusted, in addition to the variables used in Model 2. In Model 4, eGFR, HTN, DM, and DLD were adjusted, in addition to the variables used in Model 3. Subgroup analyses were conducted, based on age group (< 50 years, 50–69 years, ≥ 70 years), sex, obesity, HTN, DM, and DLD status. Sensitivity analysis was conducted by using an alternative trajectory modeling approach via group-based trajectory modeling with fixed-effects. We also performed another sensitivity analysis by repeating the analysis starting from trajectory modeling, after excluding 765 individuals who died during the first 2 years of the event accrual period.

All statistical analyses were conducted using R (version 4.3.1; R Foundation for Statistical Computing, Vienna, Austria) and SAS statistical software (version 9.4; SAS Institute Inc., Cary, NC, USA). A two-sided P value of < 0.05 was statistically significant.

Results

Clinical characteristics of the study population

Table 1 shows the baseline characteristics of the study population. The mean age of the 233,546 participants was 47.9 ± 13.3 years, and 53.6% were men. The decreasing group had the highest mean age, BMI, SBP, DBP, FBG and serum total cholesterol levels, along with the highest median serum triglyceride level, lowest serum HDL-C and eGFR, and the highest proportion of patients with HTN, DM, and DLD. The increasing group had the highest proportion of current smokers, current drinkers, and regular exercisers.

Table 1 Clinical characteristics of the study population based on the TyG index trajectories
Full size table

Risk of all-cause mortality, based on the TyG index change trajectory

During the median 8.13-year event accrual period, 7918 death events occurred. Figure 3 depicts the survival probability curves, which were adjusted for the TyG index at baseline, based on the TyG index trajectory group during the event accrual period. Survival probability was the lowest in the increasing group, followed by the stable and decreasing groups, respectively (P = 0.043).

Fig. 3
figure 3

Adjusted survival plot showing survival probability according to the TyG index trajectories, adjusted for the TyG index at baseline. Abbreviations: TyG, triglyceride-glucose

Full size image

Table 2 presents the results of the Cox proportional hazards regression analysis. In the multivariable models, the increasing group exhibited a significantly higher risk of all-cause mortality than that in the stable group. The adjusted HR (95% CI) for all-cause mortality in the increasing group was 1.06 (1.00–1.12) in Model 1, 1.17 (1.11–1.24) in Model 2, 1.14 (1.08–1.21) in Model 3, and 1.09 (1.03–1.15) in Model 4. The multivariable models did not reveal significant differences in mortality risk between the stable and decreasing groups.

Table 2 Cox proportional hazard regression analysis for all-cause mortality by trajectory groups of TyG index changes
Full size table

Supplementary Table 2 presents the results of the subgroup analysis for all-cause mortality. The fully adjusted model revealed that the increasing TyG index trajectory group had a significantly higher risk of all-cause death in the subgroups of individuals aged < 50 years, men, obesity, HTN, DM, and DLD.

Risk of CVD mortality, based on the TyG index change trajectory

A total of 651 CVD deaths and 7267 non-CVD deaths occurred during the event accrual period. The cumulative incidence rates of CVD death, adjusted for the TyG index at baseline, are depicted, based on the three TyG index trajectory groups, in Fig. 4. The increasing group exhibited a higher cumulative CVD death rate than that of the other groups, with a trend toward significance (P = 0.063). Table 3 presents the risk of CVD mortality, based on the TyG index trajectories, using competing risk analysis. In the multivariable models, the adjusted HR (95% CI) for CVD death in the increasing group, compared with the stable group, was 1.20 (0.99–1.46) in Model 1, 1.33 (1.10–1.63) in Model 2, 1.30 (1.07–1.59) in Model 3, and 1.23 (1.01–1.50) in Model 4.The increasing group was also at a higher risk of non-CVD mortality than was the stable group in all multivariate models. However, no significant difference existed in CVD death in the decreasing group, compared with that of the stable group, across the multivariate models, with adjusted HRs (95% CI) of 1.03 (0.85–1.24) in Model 1, 1.31 (0.84–1.26) in Model 2, 1.02 (0.83–1.24) in Model 3, and 1.02 (0.84–1.25) in Model 4. Figure 5a and b present the forest plots of the results of the subgroup analysis for CVD mortality, considering non-CVD death as a competing risk. The increasing TyG index trajectory group had a significantly higher risk of CVD death in the subgroups of age 50–69 years, obesity, DM, and non-DLD (Fig. 5a). Additionally, the risk of non-CVD death was significantly higher in the increasing TyG index trajectory group than in the stable group for the subgroups of age < 50 years, men, HTN, DM, and DLD (Fig. 5b).

Fig. 4
figure 4

Adjusted survival plot showing cumulative CVD death-free probability according to the TyG index trajectories, adjusted for the TyG index at baseline. Abbreviations: TyG, triglyceride-glucose; CVD, cardiovascular disease

Full size image
Fig. 5
figure 5

Forest plots of the results of the subgroup analysis for CVD mortality (a), considering non-CVD mortality (b) as a competing risk. Abbreviation: CVD, cardiovascular disease

Full size image
Table 3 Competing risk analysis for CVD mortality by trajectory groups of TyG index changes
Full size table

Sensitivity analysis

Supplementary Table 3 presents the distribution of individuals across trajectory groups, identified by two different modeling approaches: Gaussian finite mixture modeling and group-based trajectory modeling using fixed-effects modeling. All individuals previously classified in the stable group remained in the stable group, based on group-based trajectory modeling with fixed-effects modeling. However, 26.3% of individuals originally in the increasing group were reassigned to the stable group in the new classification, whereas 54.3% of individuals in the decreasing group were reclassified into the stable group. When analyzed by using an alternative modeling approach, the baseline characteristics of the study population yielded results consistent with those of the main analysis (Supplementary Table 4). The risk of all-cause mortality was significantly higher in the increasing group than in the stable group. By contrast, the all-cause mortality risk was not significantly different between the decreasing group and the stable group (Supplementary Table 5). The risk of CVD mortality and non-CVD mortality was similarly significantly higher in the increasing group than in the stable group (Supplementary Table 6).

After excluding individuals who died within the first 2 years of the event accrual period, the risk of all-cause mortality remained significantly higher in the increasing group than in the stable group (Supplementary Table 7). A significant association between the increasing group and a higher risk of both CVD mortality and non-CVD mortality, compared with the stable group, was evident in Model 2 and Model 3. However, the significance was attenuated in Model 4 (Supplementary Table 8).

Discussion

Our study unveiled the potential of TyG index changes as a predictive tool for mortality. We found that the increasing TyG index changes group had a higher risk of all-cause and CVD-specific mortality than did the stable group, whereas the risk of CVD death was similar between the decreasing group and stable group.

The influence of the TyG index on adverse health outcomes has garnered attention as a valuable tool for assessing IR. Some studies suggest that the TyG index is linked to adverse all-cause and CVD outcomes in the general population, as well as in high-risk patients, whereas other studies report contradictory results [18,19,20,21,22,23,24,25]. For example, a study of 3,524,459 Chinese adults indicated a reverse L-shaped association between the TyG index and CVD mortality [23], but a meta-analysis demonstrated no association between the TyG index and mortality [25]. These discrepancies highlight the need for further research and are likely caused by limitations such as small sample sizes, short follow-up periods, and differences in population demographics and locations.

Our study utilized a large cohort with an extended follow-up period, allowing for a more robust analysis of how TyG index changes influence mortality. By categorizing participants, based on increasing, stable, or decreasing TyG index trajectories, we better understood the dynamic nature of IR and its relationship with mortality risk. This approach addresses the shortcomings of previous studies that relied solely on baseline IR data and single-point measurements [26, 27]. IR, reflected in the components of the TyG index (i.e., glucose and triglycerides), tends to increase with age because of changes in body fat, insulin signaling, and lipid metabolism [28]; therefore, monitoring the TyG index over time is critical. Glucose variability and rising plasma triglycerides, also associated with CVD risk, further underscore the need to consider these temporal changes [29, 30].

Recent studies have demonstrated the value of tracking TyG index changes over time. For instance, in patients with type 2 DM, the baseline TyG index and its trajectories were both associated with major adverse cardiovascular events [31]. Xu et al. [32] similarly identified three distinct TyG trajectories (i.e., low, moderate, and high) in a younger population, which offered valuable insights into cardiovascular risk. Building on these findings, our study included a broader population across various age groups, making the results more generalizable. By examining these dynamic changes in the TyG index, we better understood how these fluctuations affect long-term cardiovascular and all-cause mortality, thereby enhancing risk stratification and intervention strategies.

Our subgroup analysis revealed higher all-cause mortality in individuals younger than 50, men, and individuals with obesity, HTN, DM, or DLD in the increasing TyG group. Additionally, this group showed higher CVD mortality in individuals aged 50–69 years, individuals with obesity and DM, and individuals without DLD. These findings underscore the importance of monitoring TyG index changes, especially in high-risk populations. In particular, some types of antidiabetic medications such as biguanides and thiazolidinediones can influence serum glucose and insulin levels [33]; therefore, serial monitoring of TyG index changes may be more reliable than monitoring HOMA-IR changes in patients with DM. Further research is needed to investigate the underlying causal mechanisms. Our results align with those of the Global Burden of Disease Study 2019, and highlight the rising CVD burden in the young population [34]. The increasing prevalence of DM and obesity in adults younger than 45 years contributes to this trend. Previous studies also demonstrate that persistently elevated TyG index levels in young adulthood are associated with increased CVD and mortality risks later in life [32, 35], which reinforces our findings.

The mechanism linking TyG index changes with increased mortality remains unclear, but several possible explanations have been proposed. First, higher insulin and insulin-like growth factor I levels may be secreted in response to the gradual rise in IR, leading to cellular proliferation and reduced apoptosis, which could contribute to increased carcinogenicity [36]. Second, the gradual time-dependent increase in the TyG index may lead to a greater accumulation of oxidative stress due to hypertriglyceridemia, thereby causing endothelial dysfunction and contributing to the initiation and progression of atherosclerosis [37]. Finally, prolonged exposure to hyperglycemia may promote the glycation of platelet proteins, enhance platelet reactivity and potentially increase the risk of CVD death [38].

To our knowledge, this study is the first to explore the association between long-term TyG index trajectories and all-cause and CVD mortality in a large national cohort. By leveraging a robust sample size and linking mortality data with official records from the Korea National Statistical Office, we captured dynamic changes in IR over time, providing a more comprehensive understanding of its impact on mortality risk.

Several limitations should be noted in this study. First, although we adjusted for well-known confounders, a causal relationship between TyG index trajectories and mortality could not be confirmed. Future research should consider large-scale randomized controlled trials or Mendelian randomization studies to establish causality.

Second, although TyG index changes were modeled as a time-dependent variable, other factors that also fluctuate over time such as smoking habit, alcohol consumption, and exercise were not included as time-dependent covariates. Owing to the small proportion of changes in these lifestyle factors and modeling complexity, we used only baseline values. Additionally, missing data prevented us from adjusting for socioeconomic status, potentially allowing confounding effects [39], despite adjusting for well-known risk factors for CVD mortality [35, 40, 41]. Third, our cohort likely reflected a healthier population because the participants completed three consecutive health screenings. The lower CVD mortality rate may also be partially explained by the more specific definition of CVD used in our study than that used by the 2009–2010 Korea National Health Screening Program cohort [42]. Furthermore, because this study included only Korean participants, external validation in other ethnicities is necessary to generalize the findings.

Conclusion

For individuals with a set TyG index at baseline, an increase in TyG index levels during follow-up was independently associated with a higher risk of all-cause and CVD mortality. These findings underscore the importance of serial monitoring of TyG index changes and prioritizing strategies to reduce IR to lower mortality risk.

Availability of data and materials

The data analyzed in this study cannot be shared by the authors because the data are owned by the Korean National Health Insurance Service (NHIS). Researchers can apply for access to the data via the NHIS website (https://nhiss.nhis.or.kr). Comprehensive details on the process and a provision guide are available at http://nhiss.nhis.or.kr/bd/ab/bdaba000eng.do. No datasets were generated or analysed during the current study.

Abbreviations

CVD:

Cardiovascular disease

IR:

Insulin resistance

HOMA-IR:

Homeostatic model assessment for insulin resistance

TyG:

Triglyceride-glucose

NHIS-NSC:

National Health Insurance Service-National Sample Cohort

ICD-10:

International Classification of Diseases, tenth revision

FBG:

Fasting blood glucose

BMI:

Body mass index

SBP:

Systolic blood pressure

DBP:

Diastolic blood pressure

HDL-C:

High-density lipoprotein cholesterol

eGFR:

Estimated glomerular filtration rate

DM:

Diabetes mellitus

HTN:

Hypertension

DLD:

Dyslipidemia

HR:

Hazard ratio

CI:

Confidence interval

References

  1. Mensah GA, Fuster V, Murray CJL, Roth GA. Global burden of cardiovascular diseases and risks, 1990–2022. J Am Coll Cardiol. 2023;82(25):2350–473.

    Article  PubMed  PubMed Central  Google Scholar 

  2. GBD 2021 Causes of Death Collaborators. Global burden of 288 causes of death and life expectancy decomposition in 204 countries and territories and 811 subnational locations, 1990–2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet. 2024;403(10440):2100–32.

    Article  Google Scholar 

  3. Adeva-Andany MM, Martínez-Rodríguez J, González-Lucán M, Fernández-Fernández C, Castro-Quintela E. Insulin resistance is a cardiovascular risk factor in humans. Diabetes Metab Syndr. 2019;13(2):1449–55.

    Article  PubMed  Google Scholar 

  4. Fazio S, Mercurio V, Tibullo L, Fazio V, Affuso F. Insulin resistance/hyperinsulinemia: an important cardiovascular risk factor that has long been underestimated. Front Cardiovasc Med. 2024;11:1380506.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. 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.

    Article  PubMed  Google Scholar 

  6. Son DH, Lee HS, Lee YJ, Lee JH, Han JH. Comparison of triglyceride-glucose index and HOMA-IR for predicting prevalence and incidence of metabolic syndrome. Nutr Metab Cardiovasc Dis. 2022;32(3):596–604.

    Article  CAS  PubMed  Google Scholar 

  7. Park HM, Lee HS, Lee YJ, Lee JH. The triglyceride-glucose index is a more powerful surrogate marker for predicting the prevalence and incidence of type 2 diabetes mellitus than the homeostatic model assessment of insulin resistance. Diabetes Res Clin Pract. 2021;180:109042.

    Article  CAS  PubMed  Google Scholar 

  8. Luo P, Cao Y, Li P, Li W, Song Z, Fu Z, Zhou H, Yi X, Zhu L, Zhu S. TyG index performs better than HOMA-IR in Chinese type 2 diabetes mellitus with a BMI < 35 kg/m2: A hyperglycemic Clamp Validated Study. Medicina. 2022;58(7):876–900.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Wang S, Shi J, Peng Y, Fang Q, Mu Q, Gu W, Hong J, Zhang Y, Wang W. 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.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Petersen NH, Kodali S, Meng C, Li F, Nguyen CK, Peshwe KU, Strander S, Silverman A, Kimmel A, Wang A, et al. Blood pressure trajectory groups and outcome after endovascular thrombectomy: a multicenter study. Stroke. 2022;53(4):1216–25.

    Article  CAS  PubMed  Google Scholar 

  11. Liu J, Sui X, Lavie CJ, Zhou H, Park YM, Cai B, Liu J, Blair SN. Effects of cardiorespiratory fitness on blood pressure trajectory with aging in a cohort of healthy men. J Am Coll Cardiol. 2014;64(12):1245–53.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Lee JH, Jeon S, Joung B, Lee HS, Kwon YJ. Associations of homeostatic model assessment for insulin resistance trajectories with cardiovascular disease incidence and mortality. Arterioscler Thromb Vasc Biol. 2023;43(9):1719–28.

    Article  CAS  PubMed  Google Scholar 

  13. Wang A, Tian X, Zuo Y, Chen S, Meng X, Wu S, Wang Y. Change in triglyceride-glucose index predicts the risk of cardiovascular disease in the general population: a prospective cohort study. Cardiovasc Diabetol. 2021;20(1):113.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Kim BY, Kang SM, Kang JH, Kang SY, Kim KK, Kim KB, Kim B, Kim SJ, Kim YH, Kim JH, et al. 2020 Korean Society for the Study of obesity guidelines for the management of obesity in Korea. J Obes Metab Syndr. 2021;30(2):81–92.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130(6):461–70.

    Article  CAS  PubMed  Google Scholar 

  16. Scrucca L, Fop M, Murphy TB, Raftery AE. mclust 5: clustering, classification and density estimation using Gaussian finite mixture models. R J. 2016;8(1):289–317.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Zhang X, Zhang MJ. SAS macros for estimation of direct adjusted cumulative incidence curves under proportional subdistribution hazards models. Comput Methods Programs Biomed. 2011;101(1):87–93.

    Article  PubMed  Google Scholar 

  18. Irace C, Carallo C, Scavelli FB, De Franceschi MS, Esposito T, Tripolino C, Gnasso A. Markers of insulin resistance and carotid atherosclerosis. A comparison of the homeostasis model assessment and triglyceride glucose index. Int J Clin Pract. 2013;67(7):665–72.

    Article  CAS  PubMed  Google Scholar 

  19. Sánchez-Íñigo L, Navarro-González D, Fernández-Montero A, Pastrana-Delgado J, Martínez JA. The TyG index may predict the development of cardiovascular events. Eur J Clin Invest. 2016;46(2):189–97.

    Article  PubMed  Google Scholar 

  20. Li J, Ren L, Chang C, Luo L. Triglyceride-glucose index predicts adverse events in patients with Acute Coronary Syndrome: a Meta-Analysis of Cohort Studies. Horm Metab Res. 2021;53(9):594–601.

    Article  CAS  PubMed  Google Scholar 

  21. Ding X, Wang X, Wu J, Zhang M, Cui M. Triglyceride-glucose index and the incidence of atherosclerotic cardiovascular diseases: a meta-analysis of cohort studies. Cardiovasc Diabetol. 2021;20(1):76.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Barzegar N, Tohidi M, Hasheminia M, Azizi F, Hadaegh F. The impact of triglyceride-glucose index on incident cardiovascular events during 16 years of follow-up: Tehran Lipid and Glucose Study. Cardiovasc Diabetol. 2020;19(1):155.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. He G, Zhang Z, Wang C, Wang W, Bai X, He L, Chen S, Li G, Yang Y, Zhang X, et al. Association of the triglyceride–glucose index with all-cause and cause-specific mortality: a population-based cohort study of 3.5 million adults in China. Lancet Reg Health West Pac. 2024;49:101135.

    PubMed  PubMed Central  Google Scholar 

  24. Liu C, Liang D, Xiao K, Xie L. Association between the triglyceride-glucose index and all-cause and CVD mortality in the young population with diabetes. Cardiovasc Diabetol. 2024;23(1):171.

    Article  PubMed  PubMed Central  Google Scholar 

  25. Liu X, Tan Z, Huang Y, Zhao H, Liu M, Yu P, Ma J, Zhao Y, Zhu W, Wang J. Relationship between the triglyceride-glucose index and risk of cardiovascular diseases and mortality in the general population: a systematic review and meta-analysis. Cardiovasc Diabetol. 2022;21(1):124.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Ormazabal V, Nair S, Elfeky O, Aguayo C, Salomon C, Zuñiga FA. Association between insulin resistance and the development of cardiovascular disease. Cardiovasc Diabetol. 2018;17(1):122.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Meigs JB, Mittleman MA, Nathan DM, Tofler GH, Singer DE, Murphy-Sheehy PM, Lipinska I, D’Agostino RB, Wilson PW. Hyperinsulinemia, hyperglycemia, and impaired hemostasis: the Framingham Offspring Study. JAMA. 2000;283(2):221–8.

    Article  CAS  PubMed  Google Scholar 

  28. Kolb H, Kempf K, Martin S. Insulin and aging—a disappointing relationship. Front Endocrinol. 2023;14:1261298.

    Article  Google Scholar 

  29. Li F, Zhang L, Shen Y, Liu HH, Zhang ZY, Hu G, Wang RX. Higher glucose fluctuation is associated with a higher risk of cardiovascular disease: insights from pooled results among patients with diabetes. J Diabetes. 2023;15(5):368–81.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Spitler KM, Davies BSJ. Aging and plasma triglyceride metabolism. J Lipid Res. 2020;61(8):1161–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Xu W, Zhao H, Gao L, Guo L, Liu J, Li H, Sun J, Xing A, Chen S, Wu S, et al. Association of long-term triglyceride-glucose index level and change with the risk of cardiometabolic diseases. Front Endocrinol. 2023;14:1148203.

    Article  Google Scholar 

  32. Xu X, Huang R, Lin Y, Guo Y, Xiong Z, Zhong X, Ye X, Li M, Zhuang X, Liao X. High triglyceride-glucose index in young adulthood is associated with incident cardiovascular disease and mortality in later life: insight from the CARDIA study. Cardiovasc Diabetol. 2022;21(1):155.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Bailey CJ. Treating insulin resistance in type 2 diabetes with metformin and thiazolidinediones. Diabetes Obes Metab. 2005;7(6):675–91.

    Article  CAS  PubMed  Google Scholar 

  34. Sun J, Qiao Y, Zhao M, Magnussen CG, Xi B. Global, regional, and national burden of cardiovascular diseases in youths and young adults aged 15–39 years in 204 countries/territories, 1990–2019: a systematic analysis of Global Burden of Disease Study 2019. BMC Med. 2023;21(1):222.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Aggarwal R, Yeh RW, Joynt Maddox KE, Wadhera RK. Cardiovascular risk factor prevalence, treatment, and control in us adults aged 20 to 44 years, 2009 to March 2020. JAMA. 2023;329(11):899–909.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Gallagher EJ, LeRoith D. The proliferating role of insulin and insulin-like growth factors in cancer. Trends Endocrinol Metab. 2010;21(10):610–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Bae JH, Bassenge E, Kim KB, Kim YN, Kim KS, Lee HJ, Moon KC, Lee MS, Park KY, Schwemmer M. Postprandial hypertriglyceridemia impairs endothelial function by enhanced oxidant stress. Atherosclerosis. 2001;155(2):517–23.

    Article  CAS  PubMed  Google Scholar 

  38. Schneider DJ. Factors contributing to increased platelet reactivity in people with diabetes. Diabetes Care. 2009;32(4):525–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  39. Stringhini S, Carmeli C, Jokela M, Avendaño M, Muennig P, Guida F, Ricceri F, d’Errico A, Barros H, Bochud M, et al. Socioeconomic status and the 25 × 25 risk factors as determinants of premature mortality: a multicohort study and meta-analysis of 1·7 million men and women. Lancet. 2017;389(10075):1229–37.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Cui C, Liu L, Zhang T, Fang L, Mo Z, Qi Y, Zheng J, Wang Z, Xu H, Yan H, et al. Triglyceride-glucose index, renal function and cardiovascular disease: a national cohort study. Cardiovasc Diabetol. 2023;22(1):325.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Magnussen C, Ojeda FM, Leong DP, Alegre-Diaz J, Amouyel P, Aviles-Santa L, De Bacquer D, Ballantyne CM, Bernabé-Ortiz A, Bobak M, et al. Global effect of modifiable risk factors on cardiovascular disease and mortality. N Engl J Med. 2023;389(14):1273–85.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Yi SW, An SJ, Park HB, Yi JJ, Ohrr H. Association between low-density lipoprotein cholesterol and cardiovascular mortality in statin non-users: a prospective cohort study in 14.9 million Korean adults. Int J Epidemiol. 2022;51(4):1178–89.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Nothing to declare.

Funding

This study was supported by the Yonsei University for Academic Research (Grant Number: 6-2020-0143) and a National Research Foundation of Korea (NRF) grant funded by the Korean government (Grant Number RS-2024-00354524).

Author information

Author notes

  1. Hye Sun Lee and Ji-Won Lee have contributed equally to this work.

Authors and Affiliations

  1. Department of Family Medicine, Nowon Eulji Medical Center, Eulji University School of Medicine, Seoul, 01830, Republic of Korea

    Jun-Hyuk Lee MD, PhD

  2. Department of Research Affairs, Biostatistics Collaboration Unit, Yonsei University College of Medicine, Seoul, 03277, Republic of Korea

    Soyoung Jeon PhD & Hye Sun Lee PhD

  3. Department of Family Medicine, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemun-gu, Seoul, 03722, Republic of Korea

    Ji-Won Lee MD, PhD

  4. Institute for Innovation in Digital Healthcare, Yonsei University, Seoul, 03722, Republic of Korea

    Ji-Won Lee MD, PhD

Authors
  1. Jun-Hyuk Lee MD, PhD
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Soyoung Jeon PhD
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Hye Sun Lee PhD
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Ji-Won Lee MD, PhD
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

All authors contributed substantially to the completion of this study. JHL was responsible for conceptualization. JHL, SJ, and HSL were responsible for data analysis. JHL was responsible for writing the manuscript. JHL, SJ, and JWL extracted and collated the data. HSL and JWL were jointly responsible for the design of the research and editing the manuscript. JWL was responsible for funding acquisition. All authors have read and approved the final version of the manuscript.

Corresponding authors

Correspondence to Hye Sun Lee PhD or Ji-Won Lee MD, PhD.

Ethics declarations

Ethics approval and consent to participate

This study was conducted in accordance with the STROBE Statement. The Institutional Review Board of Eulji University Hospital (Seoul, Republic of Korea) approved the study (Approval No. 2023-12-018). The requirement for informed consent was waived because we used anonymized data provided by the National Health Insurance Service-National Sample Cohort database, based on the Personal Data Protection Act guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lee, JH., Jeon, S., Lee, H.S. et al. Trajectories of triglyceride-glucose index changes and their association with all-cause and cardiovascular mortality: a competing risk analysis. Cardiovasc Diabetol 23, 364 (2024). https://doi.org/10.1186/s12933-024-02457-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12933-024-02457-y

Keywords

Cardiovascular Diabetology

ISSN: 1475-2840