Skip to main content
Download PDF
Download ePub

The association between index-year, average, and variability of the triglyceride-glucose index with health outcomes: more than a decade of follow-up in Tehran lipid and glucose study

Cardiovascular Diabetology volume 23, Article number: 321 (2024) Cite this article

Abstract

Background

The association between baseline triglyceride glucose index (TyG index) and incident non-communicable diseases, mainly in Asian populations, has been reported. In the current study, we aimed to evaluate the association between index-year, average, and visit-to-visit variability (VVV) of the TyG index with incident type 2 diabetes mellitus (T2DM), hypertension, cardiovascular disease (CVD), and all-cause mortality among the Iranian population.

Methods

The study population included 5220 participants (2195 men) aged ≥ 30 years. TyG index was calculated as Ln (fasting triglycerides (mg/dL) × fasting plasma glucose (mg/dL)/2). Average values of the TyG index and also VVV (assessed by the standard deviation (SD) and variability independent of mean) were derived during the exposure period from 2002 to 2011 (index-year). Multivariable Cox proportional hazards regression models were used to estimate the hazard ratio (HR) and 95% confidence interval (CI) of the TyG index for incident different health outcomes.

Results

During more than 6 years of follow-up after the index year, 290, 560, 361, and 280 events of T2DM, hypertension, CVD, and all-cause mortality occurred. 1-SD increase in the TyG index values at the index-year was independently associated with the incident T2DM [HR (95% CI) 2.50 (2.13–2.93)]; the corresponding values for the average of TyG index were 2.37 (2.03–2.76), 1.12 (0.99–1.26, pvalue = 0.05), 1.18 (1.01–1.36), and 1.29 (1.08–1.53) for incident T2DM, hypertension, CVD, and all-cause mortality, respectively. Compared to the first tertile, tertile 3 of VVV of the TyG index was independently associated with incident hypertension [1.33 (1.07–1.64), Ptrend <0.01]. Likewise, a 1-SD increase in VVV of the TyG index was associated with an 11% excess risk of incident hypertension [1.11 (1.02–1.21)]. However, no association was found between the VVV of the TyG index and other outcomes. Moreover, the impact of index-year and average values of the TyG index was more prominent among women regarding incident CVD (P for interactions < 0.05).

Conclusion

Although the higher TyG index at index-year and its VVV were only associated with the incident T2DM and hypertension, respectively, its average value was capable of capturing the risk for all of the health outcomes.

Introduction

The pandemic of type 2 diabetes (T2DM), hypertension, and cardiovascular disease (CVD), serious non-communicable diseases (NCDs) imposes a significant burden on global health, particularly in the Middle East and North Africa (MENA) region [1,2,3]. According to the International Diabetes Federation (IDF), in 2019, 16.2% of all-cause mortalities that occurred in the MENA region were attributed to diabetes, which was the highest proportion of diabetes-related mortality worldwide [4]. Moreover, the Global Burden of Disease (GBD) has disclosed that from 1990 to 2019, the incidence and mortality of CVD in this region have increased by 133.45% and 78.34%, respectively [5]. During the same period, the prevalence of hypertension and hypertensive heart disease in the MENA region increased by about 7% and 8%, respectively [3, 6]. The majority of CVD’s catastrophic burden is driven by five modifiable cardiometabolic risk factors, including diabetes, systolic blood pressure (SBP), body mass index (BMI), low-density lipoprotein cholesterol (LDL-C), and tobacco smoking [7].

Previously, a large number of studies have explored the association between cardiometabolic risk factors and aforementioned health outcomes by snapshot measurements or longitudinal trajectory analysis [8, 9]. Although these studies have provided valuable information, previous common methods would not consider the impact of visit-to-visit variability (VVV) on cardiometabolic risk factors. In recent years, VVV of cardiometabolic risk factors, irrespective of their mean values, has been acknowledged by researchers as a novel risk factor for T2DM, hypertension, CVD, and all-cause mortality [10, 11]. In accordance with this context, standard deviation (SD) has been reported as a reliable statistical variability index [12].

The close relationship between T2DM, hypertension, and CVD has led to the “common soil” hypothesis, which postulated shared genetic and environmental antecedents [13]. Accordingly, numerous studies have pointed to insulin resistance (IR) as one of the most critical potential antecedents [14, 15]. Moreover, some epidemiological and Mendelian randomization studies have advocated the independent association between IR and CVD [16,17,18]. Hyperinsulinaemic-euglycemic clamp and Homoeostasis model assessment of insulin resistance (HOMA-IR) are the most recommended validated tools of IR evaluation; however, a dramatic gap exists between clinical practice and their utilization due to complexity and costs [19]. Triglyceride-glucose index (TyG index), the logarithmized product of fasting plasma glucose (FPG) and triglycerides (TG), has emerged as an affordable and simple surrogate of IR [20].

In a prospective cohort study including 141,243 participants from 5 continents, the TyG index was significantly associated with mortality and CVD events [21]. While the association between TyG index values and CVD, hypertension, and T2DM has been studied extensively, very few studies have evaluated the association between TyG index variability and mentioned outcomes; principally were carried out on Asian populations [22, 23].

Despite the well-established role of ethnicity and the great burden of aforementioned NCDs, there is a lack of evidence regarding the TyG index in the MENA region [24]. Previously, we have found that in contrast to TG, FPG variability is significantly associated with the risk of CVD [25, 26]. In the current study, we aimed to extend the previous evidence by evaluating the predictive capacity of the TyG index at index-year, average value, and its variability for incident T2DM, hypertension, CVD, and all-cause mortality to the best of our knowledge for the first time in the MENA region.

Method and materials

Study design and sample

This is an observational prospective study using data from the Tehran Lipid and Glucose Study (TLGS), a large-scale community-based study initially designed in 1997 with the primary aim of preventing or postponing the development of NCDs, risk factors, and outcomes. In February 1999, TLGS was started with a cross-sectional study that included 15,005 participants over 3 years of age, residents in distinct-13 of Tehran, who were chosen by the multistage cluster random sampling method (phase I). Subsequently, phase II was prospectively started in 2002, with an additional 3550 participants and 3-year intervals of data collection, which has continued until to date (phase III: 2005–2008, Phase IV: 2008–2011, phase V: 2012–2015, phase VI: 2015–2018, phase VII: 2018–2022). The rationale and methodology of TLGS have been clarified in detail elsewhere [27].

In the current study, 5220 participants aged ≥ 30 years who entered phases II and participated in consecutive phases III and IV of TLGS (exposure period) were initially selected (men = 2195); we considered phase IV as the index-year in our data analysis. After excluding those with missing data of TyG components and confounders (considering overlap features between numbers), 4496 participants remained. Then, in order to evaluate the incidence of T2DM, hypertension, and CVD, 830, 2015, and 616 participants with prevalent mentioned diseases were excluded, respectively (Fig. 1). Ethical approval of this study has been certified by the central institutional review board of the Research Institute of Endocrine Science, Shahid Beheshti University of Medical Science, Tehran, Iran. All participants have signed written consent and outlined principles in the Declaration of Helsinki and similar ethical standards have been respected in this study.

Fig. 1
figure 1

Flowchart and timeline of the study. A Flowchart of study population, B timeline of the study. T2DM, type 2 diabetes mellitus; CVD, cardiovascular disease. *Either missing data regarding triglyceride/fasting plasma glucose during the exposure period or covariates at the index-year

Full size image

Clinical and laboratory measurements and calculations

Each participant was interviewed individually by a trained interviewer to obtain past medical history and demographic information using pretested questionnaires. Anthropometrical indices were measured as explained previously [27]. BMI was calculated as weight (kg) divided by height squared (m2). Blood pressure was measured twice with 5-min intervals in a sitting position, after at least 15 min of rest, using a standardized mercury sphygmomanometer (calibrated by the Institute of Standards and Industrial Researches), and the mean values were documented for SBP and diastolic blood pressure (DBP).

After 12–14 h of overnight fasting, venous blood samples were collected between 7:00 A.M.–9:00 A.M. for laboratory measurements; all were analyzed in the TLGS research laboratory on the collection date. FPG and TG were assayed using the enzymatic colorimetric method with glucose oxidase, and glycerol phosphate oxidase, respectively. An oral glucose tolerance test was performed, as explained previously [28]. Estimate Glomerular Filtration Rate (eGFR) was calculated by the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation [29]. Low-density lipoprotein cholesterol (LDL-C) was calculated using a modified Friedewald formula [30]. All analyses were performed using Pars Azmon kits (Pars Azmon Inc., Tehran, Iran) and Selectra 2 auto-analyzer (Vital Scientific, Spankeren, The Netherlands). Regarding FPG, both inter- and intra-assay coefficient variation (CV) were < 2.3%, and for TG, intra- and inter-assay CVs were 1.6 and 0.6%, respectively [27].

Definition of terms

In our primary analysis for different independent variables of the TyG index, we defined T2DM as FPG ≥ 7 mmol/L (126 mg/dL) or a history of glucose-lowering medication. Specifically for incident T2DM, we also defined T2DM using 2-h post-challenge glucose (2hPG) criteria, with the cut-off point of ≥ 11 mmol/L (200 mg/dL) [31]. Individuals with missing data on 2hPG who simultaneously had fasting glucose < 5 mmol/L (90 mg/dL) were considered T2DM-free [32].

Following the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC-7), hypertension was defined as SBP ≥ 140 mmHg, DBP ≥ 90 mmHg, or any history of anti-hypertensive medication. Prevalent CVD was defined as any history of percutaneous coronary intervention, coronary artery bypass graft, acute coronary syndrome leading to coronary care unit admission, angiographic-proven coronary artery disease, or stroke events. History of MI, stroke, or SCD in male and female first-degree relatives aged < 55 and < 65 years, respectively, were considered as a family history of premature CVD (FH-CVD). Diagnosed DM in first-degree relatives was defined as a family history of DM (FH-DM). Obesity was defined as BMI ≥ 30 kg/m2. The habit of smoking any tobacco product at the time of examination, either daily or occasionally, was considered a current smoker. Educational level was categorized into three levels: < 6 years of education, 6–12 years of education, and > 12 years of education. Marital status was categorized as married and single.

Main exposures

TyG index is calculated as Ln (fasting TG (mg/dL) × FPG (mg/dL)/2) [20, 33]. Besides the TyG index at the index-year as one of the exposures, we also calculated the average of the TyG index using the mean of three consecutive assessments of the TyG index during the exposure period. Moreover, the VVV of the TyG index was evaluated by the SD method [12]. We also assessed VVV, applying variability independent of mean (VIM) as another independent variable for incident health outcomes. Accordingly, it was calculated as 100 × SD/meanβ (β is the regression coefficient of the natural logarithm of the SD on the mean (Ln SD/Ln mean) [34].

CVD outcome and mortality events

Collecting data and the manner of determining outcomes in TLGS have been explained in detail previously [27]. In summary, annual telephone follow-ups were performed by trained nurses to record any new medical event. Afterward, additional information about events of interest was provided by trained physicians through home visits, reviewing medical records, and hospital documents. Ultimately, Outcomes were determined at the discretion of the outcome committee (incorporating an endocrinologist, cardiologist, epidemiologist, internist, pathologist, and any other experts if needed), based on the collected information and a particular code was assigned to each event using ICD-10 criteria and American Heart Association classification for cardiovascular events. Notably, the outcome committee was blinded to the baseline risk factors.

Based on the electrocardiogram, biomarkers, and medical records, coronary heart disease (CHD) was defined as cases of either definite or probable myocardial infarction (MI), unstable angina, angiographic-proven CHD, and CHD-related deaths. Stroke and transient ischemic attack (TIA) were defined based on the World Health Organization (WHO) criteria [35]. CVD was defined as a composite of CHD, stroke, and cerebrovascular death.

Statistical analysis

Summary statistics for the study population are expressed as mean (± SD) values and frequencies (%) for continuous and categorical variables, respectively. Comparison of the index-year characteristics of the participants stratified by the tertile of TyG index variability was performed using the ANOVA and Chi-squared test, as appropriate.

Person-years of follow-up were calculated for each participant from the index-year to the first event. Crude incidence rates of the outcomes per 1000 person-years were also calculated, and the exposure and follow-up durations were reported as medians (interquartile range). Kaplan–Meier survival curves were plotted, and cumulative incidence rates of each health outcome across the tertiles of the index-year, average, and variability of the TyG index were compared using the log-rank test (see additional file 1). Multivariable-adjusted restricted cubic spline analysis was conducted to assess the shape of the association between index-year, average, and variability of the TyG index and each health outcome (considering model 2, see below); generally, linear associations were observed (see additional file 2).

Multivariable Cox proportional hazards regression models were used to assess the association of the TyG index with different outcomes. These associations were evaluated in age- and sex-adjusted model (Model 1) and multivariable analysis (model 2), further adjusted for BMI, LDL-C, high-density lipoprotein cholesterol, eGFR, lipid-lowering medication, anti-hypertensive medication, smoking, education, marital status, SBP (except for incident hypertension which was adjusted for average of SBP), prevalent T2DM (except incident T2DM), prevalent CVD (except for incident CVD), FH-CVD, FH-DM (only for the outcome of T2DM). Model 3 was additionally adjusted for the average of the TyG index. To minimize the possible effects of change in the confounders during the exposure period, we considered their last values (measured at the index-year) in our data analysis; a similar approach was applied in other studies [36, 37].

We also examined the association between tertiles of VVV of the TyG index with different outcomes as a categorical variable using the SD method, considering the first tertile as a reference. The trend test was performed by treating the TyG index variability tertile as an ordinal variable in the Cox regression models. The Cox regression models were also applied to assess the relation between 1-SD increase in index-year TyG index, average TyG index, and TyG index variability with different outcomes in the same models, as mentioned above. The proportionality assumption in the Cox regression models was examined using the Schoenfeld residuals test, and generally, no significant departure was found from proportionality in hazards over time. For the CVD and All-cause mortality outcomes, the event date was defined with the exact date of the incidence of the event. Regarding T2DM and hypertension, the event date was defined as the time at the follow-up visit when the diagnosis was made for the first time. We defined censoring as being lost to follow-up, death, and reaching the end of the study. Follow-up time was calculated as the difference between the time of study entrance and either event date or censoring, whichever happened first. We assessed the presence of multicollinearity by calculating the variation inflation factor (VIF) statistic in the regression models. Considering the VIF of less than 5, no multicollinearity was found between variables in different models.

In order to assess the robustness of the observed results, we conducted three additional analyses. First, we re-evaluated the association between independent variables of the TyG index and incident T2DM, considering 2hPG as another diagnostic criteria. Second, we re-evaluated the association between TyG variability and incident health outcomes, using VIM instead of SD as the method of variability assessment [34]. Third, we separately assessed the effect of 1-SD increment in independent variables of the TyG index for different health outcomes among men and women and examined the effect modification of sex in the multivariate analysis.

The statistical analyses were conducted using STATA version 17 SE (Stata Corp LP, TX, USA) and R version 4.2.1, with a two-tailed p-value of less than 0.05 considered statistically significant.

Results

Baseline characteristics

Table 1 presents the baseline characteristics of eligible participants according to the tertiles of the TyG index variability at the index-year, for all-cause mortality outcome, which comprised the highest number of participants compared to other outcomes. In total, 4487 participants (37.7% men) with a mean (± SD) age and BMI of 54.4 (± 12.3) years and 29.1 (± 9.6) kg/m2 were included, respectively. Equal to the index-year, the average of the TyG index during the exposure period was 8.9 (± 0.6). In addition, participants were divided according to the tertile of TyG index variability. Generally, compared to the lowest tertile, higher percentages of smoking and consuming anti-hypertensive or lipid-lowering medications were observed in the highest tertile. Moreover, DM, FH-DM, and CVD were more prevalent in the highest tertile. TyG index, either in index-year or average values was higher in the highest tertile; however, in the case of high-density lipoprotein cholesterol (HDL-C), it was vice versa. The pattern of index-year characteristics according to tertiles of the TyG index remained essentially unchanged, considering hypertension as the outcome (see additional file 3).

Table 1 Baseline characteristics of the eligible participants for all-cause mortality analysis; Tehran lipid and glucose study
Full size table

The association between the TyG index at index-year and average values with incident health outcomes

During more than 6 years of follow-up after the exposure period, 290, 560, 361, and 280 cases of T2DM, hypertension, CVD, and all-cause mortality, with annual incidence rates of 14.87, 48.43, 11.8, and 7.68 per 1000 person-year were observed, respectively.

In categorical analysis, compared to the first tertile, the highest tertile of the TyG index at the index-year was associated with a more than six-fold increased risk of incident T2DM [HR, 95% CI 6.23 (3.96–9.79)] (see additional file 4). Furthermore, the highest tertile of the average of the TyG index showed a significant risk of incident T2DM [6.29 (4.18–10.51)]; CVD [1.47 (1.03–2.09)]; and all-cause mortality [1.55 (1.03–2.35)], compared to the first tertile (see additional file 5). The association between index-year and its average values of the TyG index and incident T2DM did not change, using 2hPG as another diagnostic criteria (n = 2705) (see additional files 6–8).

Considering the TyG index as a continuous variable showed that a 1-SD increase in the index-year values was associated with a 150% excess risk of incident T2DM [2.50 (2.13–2.93)]. Furthermore, 1-SD increase in the average value of the TyG index during the exposure period was accompanied by 137, 12, 18, and 29% elevated risk of incident T2DM, hypertension, CVD, and all-cause mortality, respectively (all pvalues ≤ 0.05) (Fig. 2B and C).

Fig. 2
figure 2

Multivariable-adjusted HRs (95% CI) for incident health outcomes per 1-SD increase in TyG index. TyG index: triglyceride-glucose index; SD, standard deviation; CVD, cardiovascular disease; T2DM, type 2 diabetes mellitus. Values are adjusted for age, sex, body mass index, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, estimated glomerular filtration rate, lipid-lowering medication, anti-hypertensive medication, smoking, education, marital status, systolic blood pressure (SBP) (except for incident hypertension which was adjusted for average of SBP), prevalent T2DM (except incident T2DM), prevalent CVD (except for incident CVD), FH-CVD, FH-DM (only for the outcome of T2DM), average values of the TyG index (for variability analysis). *pvalue= 0.05. A 1-SD increase in the variability of the TyG index; B 1-SD increase in the average values of the TyG index; C 1-SD increase of the TyG index at index-year

Full size image

Variability of TyG index and health outcomes

Table 2 shows the hazard ratio (HR, 95% CI) of incident health outcomes in three models, according to the tertile of the TyG index variability. In model 1, which was adjusted for age and sex, the highest tertile of variability was accompanied by a 35% increased risk of CVD events (Ptrend =0.01); however, this association reached null after further adjustments for conventional risk factors. In the case of hypertension, tertile 2 and 3 of the variability of the TyG index showed 27% and 33% increased incidence risk in the full model, with a significant trend (Ptrend < 0.01). Considering the TyG index variability as a continuous variable, we observed that 1-SD increment was associated with the risk of incident hypertension [1.11 (1.02–1.21)]; no such relationship was found for other outcomes (Fig. 2A).

Table 2 Association between tertiles of the TyG variability* and incident health outcomes; Tehran lipid and glucose study
Full size table

Furthermore, we re-evaluated the association between TyG variability and incident health outcomes, considering VIM to assess the variability, instead of SD. Then, in line with SD, the only significant relationship was found between VIM and incident hypertension (see additional file 9).

Subgroup analysis

We found no significant interaction between sex and independent variables of the TyG index for incident health outcomes in terms of hypertension, T2DM, and all-cause mortality. In contrast, the association between the index-year and average values of the TyG index with CVD events was more prominent among women compared to men (P for interaction = 0.037 and 0.049, respectively) (Fig. 3).

Fig. 3
figure 3

Sex-specific multivariable-adjusted HRs (95% CI) for incident health outcomes per 1-SD increase in TyG index. TyG index: triglyceride-glucose index; SD, standard deviation; CVD, cardiovascular disease; T2DM, type 2 diabetes mellitus. Values are adjusted for age, sex, body mass index, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, estimated glomerular filtration rate, lipid-lowering medication, anti-hypertensive medication, smoking, education, marital status, systolic blood pressure (SBP) (except for incident hypertension which was adjusted for average of SBP), prevalent T2DM (except incident T2DM), prevalent CVD (except for incident CVD), family history of CVD, family history of DM (only for the outcome of T2DM), average values of the TyG index (for variability analysis). A 1-SD increase in the variability of the TyG index; B 1-SD increase in the average values of the TyG index; C 1-SD increase of the TyG index at the index-year

Full size image

Discussion

In a population-based study conducted in the metropolitan city of Tehran, we examined the association between variability and average values of the TyG index during about 6 years of the exposure period, as well as its values at the index-year, with different health outcomes. Accordingly, we found that variability of the TyG index, as assessed by SD, was significantly associated with the incidence of hypertension, independent of a large set of confounders, including its average values, whereas this relationship was not found for other outcomes. Furthermore, the average value of the TyG index, independent of its variability, was a significant predictor for all of the outcomes; the corresponding risk per 1-SD increase was about 10, 20, 30, and 100% for hypertension, CVD, all-cause mortality, and T2DM, respectively. Moreover, the value of TyG at the index-year was an independent predictor only for the incidence of T2DM. Additionally, the association between index-year and average values of the TyG index was more prominent among women, compared to men, in terms of incident CVD.

Previously, numerous studies, principally in the Asian region, have individually evaluated the association between the TyG index and some NCDs, such as CVD and hypertension, by snapshot measurements [21, 38]. In the current study, the TyG index at the index-year emerged as a significant independent predictor of T2DM, aligning with a systematic review and meta-analysis encompassing 14 studies with over 270,000 participants, which has calculated the relative risk of incident T2DM for high TyG index as 3.21 (95% CI 2.75–4.54) [39]. Furthermore, a prospective cohort study involving 141,243 participants aged 35–70 from 22 countries, including Iran, has demonstrated an association between elevated TyG index and the incidence of CVD and T2DM, particularly in low- and middle-income countries [21]. Although snapshot measurements represent valuable information, they are incapable of capturing the impact of prolonged exposure and variability of the TyG index. Additionally, according to the theory of “metabolic memory” evaluating the TyG index as an IR surrogate over a long-term period provides deeper insights regarding the history of IR and possible consequences [40, 41]. In this context, we attempted to extend the previous findings by evaluating the predictive capacity of the long-term average and variability of the TyG index for the incidence of different health outcomes.

A limited number of studies have evaluated the average values of the TyG index, primarily for incident CVD and stroke [42, 43]. A prospective study with approximately 6 years of follow-up and 44,000 participants from China has claimed that participants in the fourth quartile of the cumulative TyG index have a 39% excess risk of CVD events compared to the first quartile [43]. Moreover, a retrospective study with 12 years of follow-up and about 15,000 participants from China has found a similar association between the TyG index and incident hypertension [44]. Of note; the impact of variability was not considered in the mentioned studies. Our results indicated that the long-term average of the TyG index, irrespective of its variability, has a superior predictive capacity for incidence of health outcomes rather than a single measurement at the index-year.

According to the time-varying nature of FPG and TG, two main components of the TyG index, single measurement of this index may cause potential regression dilution bias [45]. Although some studies have evaluated the impact of change in the TyG index on the incidence of different NCDs, very limited number of studies have assessed the impact of its VVV (Table 3). The vast majority of previous studies were conducted in the Asian population, which limited the generalizability of prior findings due to the influence of ethnicity and culture [23, 46]. In the present study, we found VVV of the TyG index (assessed by SD and VIM methods) as an independent predictor only for the incidence of hypertension, irrespective of conventional risk factors and average values; however, this association was not significant regarding other health outcomes. The association between the TyG index and hypertension is declared by several studies; a systematic review and meta-analysis including eight observational studies and about 200,000 participants have claimed that the highest category of TyG index is associated with the incidence of hypertension [adjusted risk ratio: 1.53 (1.26–1.85); I2 = 54%]; however, the impact of VVV was not considered [47]. Furthermore, a cohort study from China, including approximately 18,000 participants with a mean age of 60.5 years, has noted that TyG change (assessed by subtracting baseline value from the end of the follow-up) was not associated with the incidence of hypertension, whereas VVV was not evaluated in this study [48]. On the other hand, two other studies from China have advocated the independent association between VVV (assessed by SD, coefficient of variation, and average real variability) of the TyG index and the incidence of T2DM and CVD, which contradicts our results [46, 49]. Finally, it is worth noting that the observed difference between our and previous results could be interpreted by some essential issues, such as ethical differences, the method of evaluating the VVV, and the manner of dealing with the confounders. To the best of our knowledge, this is the first report on the association between VVV of the TyG index and hypertension.

Table 3 Literature review of studies evaluated the association between TyG variability and incident health outcomes
Full size table

Our epidemiological findings regarding the association between the TyG index and hypertension could be interpreted by IR, which is accompanied by compensatory hypersecretion of insulin in the early stages. Following hyperinsulinemia, sympathetic nerves would be hyper-excited, which induces renin secretion, increased sodium reabsorption, and subsequent increase of vascular resistance and cardiac output [50]. Additionally, hyper-excited sympathetic nerves promote catecholamine release, and prolonged high plasma concentration of catecholamines is a major risk factor for vascular damage through several mechanisms [51]. In later stages of IR, endothelial cells’ production and response to the vasodilators are suppressed, rooted in persistent exposure to hyperglycemia and oxidative stress [52, 53]. Interestingly, these biologically plausible explanations are also supported by epidemiological findings of two systematic reviews and meta-analyses, which have claimed that the highest category of TyG index is associated with approximately 83% and 96% augmented risk of arterial stiffness [54, 55].

The current study has some strengths that should be noted; this is the first long-term population-based prospective study that has assessed variability and average values of the TyG index in the MENA region, which has a high burden of NCDs. This study has provided a holistic overview regarding the clinical implication of three independent variables of the TyG index for predicting the incidence of major health outcomes. However, our results should be interpreted with caution due to several limitations: First, this study was conducted among residents of the metropolitan city of Tehran, limiting our results’ generalizability. Second, although we tried to adjust a wide spectrum of conventional confounders for each health outcome specifically, we cannot refuse the possibility of some residual confounders, which is the inherent feature of observational studies. Third, hemoglobin A1C was not assessed among TLGS participants. Finally, nutritional habits and energy intake, which might affect the TyG index values, were not recorded; however, related confounders such as lipid indices and BMI were adjusted in the analysis.

Conclusion

In a population-based study, TyG values at the index year were only associated with the incidence of T2DM. Regarding the TyG average values that were assessed during about 6 years, besides hypertension, strong association was also demonstrated for incident T2DM, CVD, and all-cause mortality events. Last but not least, the VVV of the TyG index was a significant predictor of hypertension, even after considering a large set of confounders including its average values. So, we hypothesize that the average values of the TyG index, rather than its single measurement, could be implemented in clinical practice to estimate the risk of different health outcomes. Moreover, keeping the TyG index in a narrow range is recommended to prevent hypertension.

Availability of data and materials

The used datasets for analysis of the current study are available from the corresponding author upon reasonable request.

Abbreviations

TLGS:

Tehran lipid and glucose study

MENA:

Middle East and North Africa

IR:

Insulin resistance

FPG:

Fasting plasma glucose

TG:

Triglyceride

TyG index:

Triglyceride glucose index, calculated as Ln (fasting triglycerides (mg/dL) × fasting plasma glucose (mg/dL)/2)

NCD:

Non-communicable disease

T2DM:

Type 2 diabetes mellitus

2hPG:

2-Hour post-challenge glucose

CHD:

Coronary heart disease

CVD:

Cardiovascular disease

VVV:

Visit-to-visit variability

SD:

Standard deviation

VIM:

Variability independent of mean

HR:

Hazard ratio

CI:

Confidence interval

BMI:

Body mass index

SBP:

Systolic blood pressure

LDL-C:

Low-density lipoprotein cholesterol

HDL-C:

High-density lipoprotein cholesterol

eGFR:

Estimated glomerular filtration rate

FH-DM:

Family history of DM

prevalent CVD:

History of diagnosed CVD

FH-CVD:

Family history of CVD

CKD-EPI:

Chronic Kidney Disease Epidemiology Collaboration equation

References

  1. Azizi F, Hadaegh F, Hosseinpanah F, Mirmiran P, Amouzegar A, Abdi H, et al. Metabolic health in the Middle East and North Africa. Lancet Diabetes Endocrinol. 2019;7(11):866–79.

    Article  PubMed  Google Scholar 

  2. Yusuf S, Joseph P, Rangarajan S, Islam S, Mente A, Hystad P, et al. Modifiable risk factors, cardiovascular disease, and mortality in 155,722 individuals from 21 high-income, middle-income, and low-income countries (PURE): a prospective cohort study. Lancet. 2020;395(10226):795–808.

    Article  PubMed  Google Scholar 

  3. Mohammadian Khonsari N, Shahrestanaki E, Ejtahed HS, Djalalinia S, Sheidaei A, Hakak-Zargar B, et al. Long-term trends in hypertension prevalence, awareness, treatment, and control rate in the Middle East and North Africa: a systematic review and meta-analysis of 178 population-based studies. Curr Hypertens Rep. 2021;23(9):41.

    Article  PubMed  Google Scholar 

  4. Saeedi P, Salpea P, Karuranga S, Petersohn I, Malanda B, Gregg EW, et al. Mortality attributable to diabetes in 20–79 years old adults, 2019 estimates: results from the International Diabetes Federation Diabetes Atlas. Diabetes Res Clin Pract. 2020;162:108086.

    Article  PubMed  Google Scholar 

  5. Li Y, Cao G-y, Jing W-z, Liu J, Liu M. Global trends and regional differences in incidence and mortality of cardiovascular disease, 1990–2019: findings from 2019 global burden of disease study. Eur J Prev Cardiol. 2023;30(3):276–86.

    Article  PubMed  Google Scholar 

  6. Dai H, Bragazzi NL, Younis A, Zhong W, Liu X, Wu J, et al. Worldwide trends in prevalence, mortality, and disability-adjusted life years for hypertensive heart disease from 1990 to 2017. Hypertension. 2021;77(4):1223–33.

    Article  PubMed  CAS  Google Scholar 

  7. Consortium GCR. Global effect of modifiable risk factors on cardiovascular disease and mortality. N Engl J Med. 2023;389(14):1273–85.

    Article  Google Scholar 

  8. McEvoy JW, Daya N, Rahman F, Hoogeveen RC, Blumenthal RS, Shah AM, et al. Association of isolated diastolic hypertension as defined by the 2017 ACC/AHA blood pressure guideline with incident cardiovascular outcomes. JAMA. 2020;323(4):329–38.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Kan Y, Liu L, Li X, Pang J, Bi Y, Zhang L, et al. Association between distinct body mass index trajectories according to the group-based trajectory modeling and the risk of incident diabetes: a systematic review. Obes Rev. 2022;23(12):e13508.

    Article  PubMed  Google Scholar 

  10. Stevens SL, Wood S, Koshiaris C, Law K, Glasziou P, Stevens RJ, et al. Blood pressure variability and cardiovascular disease: systematic review and meta-analysis. BMJ. 2016;354:i4098.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Echouffo-Tcheugui JB, Zhao S, Brock G, Matsouaka RA, Kline D, Joseph JJ. Visit-to-visit glycemic variability and risks of cardiovascular events and all-cause mortality: the ALLHAT study. Diabetes Care. 2019;42(3):486–93.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Mena L, Pintos S, Queipo NV, Aizpúrua JA, Maestre G, Sulbarán T. A reliable index for the prognostic significance of blood pressure variability. J Hypertens. 2005;23(3):505–11.

    Article  PubMed  CAS  Google Scholar 

  13. Stern MP. Diabetes and cardiovascular disease: the “common soil” hypothesis. Diabetes. 1995;44(4):369–74.

    Article  PubMed  CAS  Google Scholar 

  14. Ceriello A, Motz E. Is oxidative stress the pathogenic mechanism underlying insulin resistance, diabetes, and cardiovascular disease? The common soil hypothesis revisited. Arterioscler Thromb Vasc Biol. 2004;24(5):816–23.

    Article  PubMed  CAS  Google Scholar 

  15. Hill MA, Yang Y, Zhang L, Sun Z, Jia G, Parrish AR, et al. Insulin resistance, cardiovascular stiffening and cardiovascular disease. Metabolism. 2021;119:154766.

    Article  PubMed  CAS  Google Scholar 

  16. Si S, Li J, Li Y, Li W, Chen X, Yuan T, et al. Causal effect of the triglyceride-glucose index and the joint exposure of higher glucose and triglyceride with extensive cardio-cerebrovascular metabolic outcomes in the UK Biobank: a mendelian randomization study. Front Cardiovasc Med. 2021;7:583473.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Liu X, Tan Z, Huang Y, Zhao H, Liu M, Yu P, et al. 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):1–17.

    Article  CAS  Google Scholar 

  18. Klarin D, Zhu QM, Emdin CA, Chaffin M, Horner S, McMillan BJ, et al. Genetic analysis in UK Biobank links insulin resistance and transendothelial migration pathways to coronary artery disease. Nat Genet. 2017;49(9):1392–7.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  19. Minh HV, Tien HA, Sinh CT, Thang DC, Chen CH, Tay JC, et al. Assessment of preferred methods to measure insulin resistance in Asian patients with hypertension. J Clin Hypertens. 2021;23(3):529–37.

    Article  CAS  Google Scholar 

  20. 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 

  21. Lopez-Jaramillo P, Gomez-Arbelaez D, Martinez-Bello D, Abat MEM, Alhabib KF, Avezum Á, et al. Association of the triglyceride glucose index as a measure of insulin resistance with mortality and cardiovascular disease in populations from five continents (PURE study): a prospective cohort study. Lancet Healthy Longev. 2023;4(1):e23-33.

    Article  PubMed  Google Scholar 

  22. Li H, Zuo Y, Qian F, Chen S, Tian X, Wang P, et al. Triglyceride-glucose index variability and incident cardiovascular disease: a prospective cohort study. Cardiovasc Diabetol. 2022;21(1):105.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Xu W, Zhao H, Gao L, Guo L, Liu J, Li H, et al. Association of long-term triglyceride-glucose index level and change with the risk of cardiometabolic diseases. Front Endocrinol (Lausanne). 2023;14:1148203.

    Article  PubMed  Google Scholar 

  24. Khan NA, Wang H, Anand S, Jin Y, Campbell NR, Pilote L, et al. Ethnicity and sex affect diabetes incidence and outcomes. Diabetes Care. 2011;34(1):96–101.

    Article  PubMed  Google Scholar 

  25. Barzegar N, Ramezankhani A, Tohidi M, Azizi F, Hadaegh F. Long-term glucose variability and incident cardiovascular diseases and all-cause mortality events in subjects with and without diabetes: Tehran lipid and glucose study. Diabetes Res Clin Pract. 2021;178:108942.

    Article  PubMed  CAS  Google Scholar 

  26. Masrouri S, Cheraghi L, Deravi N, Cheraghloo N, Tohidi M, Azizi F, et al. Mean versus variability of lipid measurements over 6 years and incident cardiovascular events: more than a decade follow-up. Front Cardiovasc Med. 2022;9:1065528.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  27. Azizi F, Ghanbarian A, Momenan AA, Hadaegh F, Mirmiran P, Hedayati M, et al. Prevention of non-communicable disease in a population in nutrition transition: Tehran lipid and glucose study phase II. Trials. 2009;10:1–15.

    Article  Google Scholar 

  28. Abdi A, Kohansal K, Khalili D, Azizi F, Hadaegh F. The difference between 2-hour post-challenge and fasting plasma glucose associates with the risk of cardiovascular disease in a normoglycemic population: the Tehran lipid and glucose study. Nutr Metab. 2024;21(1):10.

    Article  CAS  Google Scholar 

  29. Nelson RG, Grams ME, Ballew SH, Sang Y, Azizi F, Chadban SJ, et al. Development of risk prediction equations for incident chronic kidney disease. JAMA. 2019;322(21):2104–14.

    Article  PubMed  PubMed Central  Google Scholar 

  30. Chen Y, Zhang X, Pan B, Jin X, Yao H, Chen B, et al. A modified formula for calculating low-density lipoprotein cholesterol values. Lipids Health Dis. 2010;9:52.

    Article  PubMed  PubMed Central  Google Scholar 

  31. 2. Diagnosis and classification of diabetes: standards of care in diabetes-2024. Diabetes Care. 2024;47(Suppl 1):S20-42.

    Google Scholar 

  32. Bozorgmanesh M, Hadaegh F, Saadat N, Azizi F. Fasting glucose cutoff point: where does the risk terminate? Tehran lipid and glucose study. Acta Diabetol. 2012;49:341–8.

    Article  PubMed  CAS  Google Scholar 

  33. Guerrero-Romero F, Simental-Mendía LE, González-Ortiz M, Martínez-Abundis E, Ramos-Zavala MaG, Hernández-González SO, et al. The product of triglycerides and glucose, a simple measure of insulin sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp. J Clin Endocrinol Metab. 2010;95(7):3347–51.

    Article  PubMed  CAS  Google Scholar 

  34. Rothwell PM, Howard SC, Dolan E, O’Brien E, Dobson JE, Dahlöf B, et al. Prognostic significance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet. 2010;375(9718):895–905.

    Article  PubMed  Google Scholar 

  35. Aho K, Harmsen P, Hatano S, Marquardsen J, Smirnov VE, Strasser T. Cerebrovascular disease in the community: results of a WHO collaborative study. Bull World Health Organ. 1980;58(1):113.

    PubMed  PubMed Central  CAS  Google Scholar 

  36. Kim MK, Han K, Park Y-M, Kwon H-S, Kang G, Yoon K-H, et al. Associations of variability in blood pressure, glucose and cholesterol concentrations, and body mass index with mortality and cardiovascular outcomes in the general population. Circulation. 2018;138(23):2627–37.

    Article  PubMed  CAS  Google Scholar 

  37. Lee S-R, Choi Y-J, Choi E-K, Han K-D, Lee E, Cha M-J, et al. Blood pressure variability and incidence of new-onset atrial fibrillation: a nationwide population-based study. Hypertension. 2020;75(2):309–15.

    Article  PubMed  CAS  Google Scholar 

  38. Liu Q, Zhang Y, Chen S, Xiang H, Ouyang J, Liu H, et al. Association of the triglyceride-glucose index with all-cause and cardiovascular mortality in patients with cardiometabolic syndrome: a national cohort study. Cardiovasc Diabetol. 2024;23(1):80.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  39. Pranata R, Huang I, Irvan, Lim MA, Vania R. The association between triglyceride-glucose index and the incidence of type 2 diabetes mellitus—a systematic review and dose–response meta-analysis of cohort studies. Endocrine. 2021;74(2):254–62.

    Article  PubMed  CAS  Google Scholar 

  40. Testa R, Bonfigli AR, Prattichizzo F, La Sala L, De Nigris V, Ceriello A. The metabolic memory theory and the early treatment of hyperglycemia in prevention of diabetic complications. Nutrients. 2017;9(5):437.

    Article  PubMed  PubMed Central  Google Scholar 

  41. Miller RG, Orchard TJ. Understanding metabolic memory: a tale of two studies. Diabetes. 2020;69(3):291–9.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  42. Wang X, Feng B, Huang Z, Cai Z, Yu X, Chen Z, et al. Relationship of cumulative exposure to the triglyceride-glucose index with ischemic stroke: a 9-year prospective study in the Kailuan cohort. Cardiovasc Diabetol. 2022;21(1):66.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  43. Cui H, Liu Q, Wu Y, Cao L. Cumulative triglyceride-glucose index is a risk for CVD: a prospective cohort study. Cardiovasc Diabetol. 2022;21(1):22.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  44. Xin F, He S, Zhou Y, Jia X, Zhao Y, Zhao H. The triglyceride glucose index trajectory is associated with hypertension: a retrospective longitudinal cohort study. Cardiovasc Diabetol. 2023;22(1):347.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  45. Robins JM, Hernan MA, Brumback B. Marginal structural models and causal inference in epidemiology. Lww. 2000;11:550–60.

    CAS  Google Scholar 

  46. Wang X, Chen Y, Huang Z, Cai Z, Yu X, Chen Z, et al. Visit-to-visit variability in triglyceride-glucose index and diabetes: a 9-year prospective study in the Kailuan study. Front Endocrinol. 2022;13:1054741.

    Article  Google Scholar 

  47. Wang Y, Yang W, Jiang X. Association between triglyceride-glucose index and hypertension: a meta-analysis. Front Cardiovasc Med. 2021;8:644035.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  48. Wang D, Li W, Zhou M, Ma J, Guo Y, Yuan J, et al. Association of the triglyceride-glucose index variability with blood pressure and hypertension: a cohort study. QJM Int J Med. 2024;117(4):277–82.

    Article  CAS  Google Scholar 

  49. Chen F, Pan Y, Liu Z, Huang R, Wang J, Shao J, et al. Impact of visit-to-visit triglyceride-glucose index variability on the risk of cardiovascular disease in the elderly. Int J Endocrinol. 2022;2022:5125884.

    Article  PubMed  PubMed Central  Google Scholar 

  50. da Silva AA, do Carmo JM, Li X, Wang Z, Mouton AJ, Hall JE. Role of hyperinsulinemia and insulin resistance in hypertension: metabolic syndrome revisited. Can J Cardiol. 2020;36(5):671–82.

    Article  PubMed  Google Scholar 

  51. Li Z, Yu C, Han Y, Ren H, Shi W, Fu C, et al. Inhibitory effect of D1-like and D3 dopamine receptors on norepinephrine-induced proliferation in vascular smooth muscle cells. Am J Physiol Heart Circ Physiol. 2008;294(6):H2761–8.

    Article  PubMed  CAS  Google Scholar 

  52. Usui I. Common metabolic features of hypertension and type 2 diabetes. Hypertens Res. 2023;46(5):1227–33.

    Article  PubMed  CAS  Google Scholar 

  53. Jia G, Sowers JR. Hypertension in diabetes: an update of basic mechanisms and clinical disease. Hypertension. 2021;78(5):1197–205.

    Article  PubMed  CAS  Google Scholar 

  54. Liu F, Ling Q, Xie S, Xu Y, Liu M, Hu Q, et al. Association between triglyceride glucose index and arterial stiffness and coronary artery calcification: a systematic review and exposure-effect meta-analysis. Cardiovasc Diabetol. 2023;22(1):111.

    Article  PubMed  PubMed Central  Google Scholar 

  55. Sajdeya O, Beran A, Mhanna M, Alharbi A, Burmeister C, Abuhelwa Z, et al. Triglyceride glucose index for the prediction of subclinical atherosclerosis and arterial stiffness: a meta-analysis of 37,780 individuals. Curr Probl Cardiol. 2022;47(12):101390.

    Article  PubMed  Google Scholar 

  56. Chen N, Ma L-L, Zhang Y, Chu X, Dong J, Yan Y-X. Association of long-term triglyceride-glucose index patterns with the incidence of chronic kidney disease among non-diabetic population: evidence from a functional community cohort. Cardiovasc Diabetol. 2024;23(1):7.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

Download references

Acknowledgements

We sincerely wish to thank the study participants and the TLGS executive team for their passionate support.

Funding

No funding to declare related to this work.

Author information

Author notes

  1. Danial Molavizadeh and Neda Cheraghloo have contributed equally to this work.

Authors and Affiliations

  1. Prevention of Metabolic Disorders Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, P.O. Box 19395-4763, Tehran, Islamic Republic of Iran

    Danial Molavizadeh, Maryam Tohidi & Farzad Hadaegh

  2. School of Medicine, Kashan University of Medical Sciences, Kashan, Islamic Republic of Iran

    Danial Molavizadeh

  3. Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Science, Tehran, Islamic Republic of Iran

    Neda Cheraghloo

  4. Endocrine Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Islamic Republic of Iran

    Fereidoun Azizi

Authors
  1. Danial Molavizadeh
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Neda Cheraghloo
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Maryam Tohidi
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Fereidoun Azizi
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Farzad Hadaegh
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

DM, FH, MT, and NC conceived and planned the study. DM performed literature research. NC conducted the analyses. FH, DM, MT, and NC contributed to the interpretation of the data. FH and MT conducted a critical revision of the manuscript. DM developed the first draft of the manuscript. All authors have read and approved the final manuscript.

Corresponding author

Correspondence to Farzad Hadaegh.

Ethics declarations

Ethics approval and consent to participate

Ethical approval of this study has been certified by the central institutional review board of the Research Institute of Endocrine Science, Shahid Beheshti University of Medical Science, Tehran, Iran. The outlined principles in the Declaration of Helsinki have been respected in this study. All participants have signed written consent.

Consent for publication

All authors have declared their consent for publication.

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.

Electronic supplementary material

12933_2024_2387_MOESM1_ESM.tiff

Additional file 1: Kaplan-Meier survival curves according to the tertiles of the independent variables of the TyG index for incident type 2 diabetes, hypertension, cardiovascular disease, and all-cause mortality.

12933_2024_2387_MOESM2_ESM.tiff

Additional file 2: Multivariable-adjusted restricted cubic spline for the relationship between independent variables of the TyG index and incident type 2 diabetes, hypertension, cardiovascular disease, and all-cause mortality. Adjusted for: age, sex, body mass index, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, estimated glomerular filtration rate, lipid-lowering medication, anti-hypertensive medication, smoking, education, marital status, systolic blood pressure (SBP) (except for incident hypertension which was adjusted for average of SBP), prevalent T2DM (except incident T2DM), prevalent CVD (except for incident CVD), family history of CVD, and family history of DM (only for the outcome of T2DM).

Additional file 3.

Additional file 4.

Additional file 5.

Additional file 6.

Additional file 7.

Additional file 8.

Additional file 9.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Molavizadeh, D., Cheraghloo, N., Tohidi, M. et al. The association between index-year, average, and variability of the triglyceride-glucose index with health outcomes: more than a decade of follow-up in Tehran lipid and glucose study. Cardiovasc Diabetol 23, 321 (2024). https://doi.org/10.1186/s12933-024-02387-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12933-024-02387-9

Keywords

Cardiovascular Diabetology

ISSN: 1475-2840