Association between Triglyceride Glucose Index and the Risk of Peripheral Artery Disease

Background The triglyceride glucose (TyG) index has been used as a simple surrogate marker of insulin resistance, an independent predictor of atherosclerotic vascular diseases. However, few studies have investigated the relationship between the TyG index and peripheral artery disease (PAD). Methods A total of 3375 participants with comleted TyG and ankle brachial pressure index (ABPI) records were enrolled from the National Health and Nutrition Examination Survey (NHANES) 1999-2004. The TyG index was calculated as ln[triglycerides (mg/dL)×glucose (mg/dL)/2], and the presence of PAD was dened as ABPI ≤ 0.9. Results The participants were 60.1±12.8 year old and 51.3% (1730) were male. The prevelance of PAD was 7.1% (238). Compared with the reference lowest quartile of TyG index, the highest quartile was associated with 1.66-fold (odds ratio [OR], 95% condence interval [CI] 1.15-2.43; p=0.008) risk of PAD. After adjusted for sociodemographic, lifestyles, and cardiometabolic factors, the multivariate-adjusted OR and 95% CI were 1.55 (1.03-2.37; p=0.039) or participants within the highest quartile. TyG index was also independently and linearly associated with higher presence of PAD (OR 1.27 [1.02-1.56]; p=0.027). Subgroup analysis showed that the association between TyG index and the risk of PAD was still consistant across groups except for obesity. Conclusions TyG index was signicantly associated with the higher risk of PAD, which could be a marker of PAD.


Background
Cardiovascular disease (CVD) is a leading cause of morbility and mortality worldwide [1]. Peripheral artery disease (PAD), de ned as ankle-brachial pressure index (ABPI) ≤ 0.9, has been established as an independent marker for atherosclerotic vascular diseases [2] and a predictor for all-cause mortality and cardiovascular event [3]. Screening the risk factors of PAD was necessary to reduce the complications.
The triglyceride glucose (TyG) index has been regared as a reliable surrogate marker of insulin resistance (IR) [4,5], characterized by poor insulin sensitivity in the peripheral tissues [6]. There were growing evidences to show that IR contributed to vascular remodeling and incresead vascular calci cation [7,8], contributing to the risk of cardiovascular diseases and mortality [9]. Many studies have identi ed TyG index was associated with cardiovascular diseases [10,11]. However, few study has invesitgated the assocaition between the TyG index and the risk of PAD in healthy adults.
In the present study, we evaluated the association between the TyG index and the risk of PAD in genral population.

Study population
The study used data from National Health and Nutrition Examination Survey (NHANES) between the periods of 1999 to 2004, a nationwide survey conducted by the National Center of Health Statistics (NCHS). After excluding participants with missing data on triglyceride (328) and glucose (3484) from 7187 participants with ABPI records, 3375 participants were enrolled in our analysis ( Figure 1). The study was approved by the institutional review board of NCHS and all participants provided written informed consent.

Exposure variable and outcomes
The TyG index was determined as ln (triglycerides [mg/dL]×glucose [mg/dL]/2). Plasma triglyceride was measured enzymatically using Roche Modular P chemistry analyzer and fasting glucose was measured by an enzymatic method. The ABPI exam was performed by trained health technicians in the mobile examination center. Participants lie supine on the exam table during the exam. Systolic pressure is measured on the right arm (brachial artery) and both ankles (posterior tibial arteries). Systolic blood pressure is measured twice at each site for participants aged 40-59 years and once at each site for participants aged 60 years and older. The ABPI was automatically calculated by the computer system and veri ed by NCHS. The presence of PAD was de ned as any of left or right ABPI ≤ 0.9.

Covariates collection
The baseline characteristics of participants were acquired by questionnaires and examinations, including sociodemographic (gender, age, race, and educational level), lifestyle information (physical activity and smoking status), and mediations use (hypoglycemic drugs and lipid-lowering drugs). Systolic blood pressure and diastolic blood pressure were measured by physical examination. Body mass index (BMI) was calculated as body weight divided by height squared. Race was classi ed as non-Hispanic white, non-Hispanic black, Mexican American, other Hispanic or others. Education level was categorized as less than high school, high school, equivalent and college or above. Smoking status were de ned as current, past and never. Physical activity status was classi ed as vigorous, moderate and inactive. Vigorous physical activity was de ned as an activity that greatly increases the breathing or heart rate. Moderate physical activity was de ned as an activity that causes small increases in breathing. Multiple imputation using predictive mean matching (PMM) was performed for covariates with missing values.

Statistical Analysis
Continuous variables were described as the mean ± standard deviation while categorical variables were presented as numbers and proportions (percentage). Differences between groups were explored by one-way analysis of variance (ANOVA) or chi-square tests. Multivariate logistic regression models were used to estimate odds ratios (ORs) and 95% con dence interval (CI) between TyG qurtaile and PAD. Multivariate linear regression models were used to explore the risk of PAD per one-unit increasement of TyG index. Model 1 was not adjusted. Model 2 was adjusted for age, gender, and race. Model 3 was adjusted for age, gender, race, education level, BMI, smoking status, and physical activity. Model 4 was adjusted for age, gender, race, education level, BMI, smoking status, physical activity, systolic blood pressure, diastolic blood pressure, hypoglycemic drug, and lipid-lowering drug.
Subgroup analyses was performed to explore the interactions modifying the relationship. All statistical analyses were performed using R version 3.6, and P<0.05 was considered as statistically signi cant.

Results
Participants were strati ed into four groups according to their TyG quartiles. The baseline characteristics of the study population were shown in Table 1. The highest TyG quartile tended to have more percentage of male,  Subgroup analysis for the associations between TyG index and the presence of PAD was shown in Fig. 2. The association was consistent across gender, elderly and race except for BMI (P for interaction = 0.002). The positive correlation between TyG index and PAD disappeared among obesity population.

Discussion
In this study, we investigated the association between the TyG index and PAD in general U.S. adults. We found that a higher TyG index were signi cantly associated with the prevalence of PAD. And this relationship disappeared in obesity individuals.
The TyG index has been proposed as an alternative surrogate marker for insulin resistance [12,13]. Several studies demonstrated that the TyG index was positively correlated with HOMA-IR [14], and even has a better predictive value that HOMA-IR [15]. The TyG index was reported to be related to a higher risk of cardiovascular diseases and mortality [16]. In addition, some studies showed that the TyG index was signi cantly associated with the severity of coronary artery stenosis [17], artery stiffness [18] and vascular calci cation [19]. Previous study found that the TyG index was an independent predictor of peripheral artery disease complexity based on a small sample size [20]. Our results further con rmed an independent association between the TyG index and the presence of PAD, and a higher TyG index increased the risk of PAD. Besides, we found that obesity was a factor affecting the relationship between TyG and PAD. The underlying interaction [21] required further research.
The mechanism underlying the relationship could be linked to IR. IR could lead to vascular in ammation and stiffness, which contributing to atherosclerosis of periphery artery [22,23].
Some limitations existed in our study. Firstly, this was a cross-sectional study which could not infer causality.
Second, the PAD was de ned based on ABPI, which was lack of angiography examination.

Conclusions
In our study, we demonstrated that the TyG index was independently associated with the prevalence of PAD in U.S adults, which may serve as a potential predictive marker. Figure 1 The ow chart of participant selection.