Association between triglyceride glucose index and arterial stiffness in Korean adults

The study population consisted of 5989 Korean subjects who participated in a comprehensive health examination as part of a self-referred health checkup program at the Gangnam Severance Hospital Health Promotion Center from January 2008 to February 2015. The exclusion criteria of this study were: subjects with elevated triglyceride levels (> 400 mg/dl) and subjects with any history of malignancy, acute inflammatory disease, infectious disease, or renal disease. We also excluded subjects with a history of hypertension, angina, myocardial infarction, cerebrovascular accident, or an ankle–brachial index (ABI) < 0.90. Subjects taking statins or triglyceride-lowering medications (fenofibrate or omega-3) were also excluded. In addition, we excluded men and women with a history of alcohol consumption in excess of 30 and 20 g/day, respectively. After applying the exclusion criteria, a total of 3587 subjects were enrolled in our final analysis. The study protocol was approved by the Institutional Review Board of Yonsei University College of Medicine.

We measured the height and weight of each study participant and calculated the body mass index (BMI, kg/m²). Lifestyle information, personal medical history of acute and chronic illnesses, and medication history were collected using a standard questionnaire. Systolic and diastolic blood pressures (SBP, DBP) were measured by experienced technicians using an automated blood pressure (BP) monitor (HEM-7080IC; Omron Healthcare, Lake Forest, IL, USA) after a 5-min rest period with the patient’s arm placed at heart level. The diagnosis of diabetes mellitus was based on a previous history of diabetes or the ADA’s diagnostic guidelines. Hypertension was defined as systolic and/or diastolic blood pressures > 140/90 mmHg or the current use of an antihypertensive medication. Current smokers were defined as having smoked cigarettes regularly over the previous 6 months.

Blood samples were obtained from all subjects after 8 h of fasting. Samples were immediately centrifuged, and the serum samples were stored at − 70 °C until analysis. The fasting plasma glucose (FPG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TG) were determined using enzymatic methods with a Hitachi 7600–120 automated chemistry analyzer (Hitachi, Tokyo, Japan). Low-density lipoprotein cholesterol (LDL-C) was calculated according to the Friedewald formula. The TyG index was calculated as ln[fasting triglycerides (mg/dl) × fasting glucose (mg/dl)/2] [7]. The fasting serum insulin level was determined by chemiluminescence with an RIA kit (Daiichi, Japan). Insulin resistance was estimated using the homeostasis model assessment of insulin resistance (HOMA-IR) index calculated as fasting insulin (μU/mL) × fasting plasma glucose (mmol/L)/22.5.

Arterial stiffness was measured using an automatic plethysmographic instrument (VP-1000; Colin, Komaki, Japan) as previously described [17]. Briefly, electrodes were placed on both wrists, and cuffs were wrapped around both the upper arms and ankles. After simultaneous measurement of blood pressure and waveforms in all four limbs, the time interval between the brachial and ankle waveforms (ΔTba) was determined. The distance between the brachium and the ankle (La − Lb) was estimated automatically according to the subject’s height. After these data were collected, baPWV was calculated using the following equation: baPWV = (La − Lb)/ΔTba (in cm/s). Both baPWV values were measured after allowing the patient to rest in a supine position for at least 5 min. The mean of the right and left baPWV was used as a marker of arterial stiffness.

Continuous variables with normal distributions were expressed as the mean ± standard deviation (SD), whereas continuous variables with skewed distributions were expressed as the medians with interquartile ranges and were log transformed for analysis. Intergroup comparisons were performed using ANOVA. Chi square tests were used to compare categorical variables with percentages. Age-adjusted baPWV means and standard errors were calculated using analysis of covariance (ANCOVA) according to TyG quartiles. The relationships between the baPWV and various clinical parameters were examined using Pearson’s correlation. The odds ratios (OR) and corresponding 95% confidence intervals (CI) for high PWVs were estimated by multivariate logistic regression analysis after adjusting for confounding variables across TyG index and HOMA-IR quartiles. Covariates in the multivariable model, which were chosen for their clinical importance, as well as statistical significance, were age, SBP, BMI, LDL-C, HDL-C, diabetes mellitus, and menopause in women. Because an absolute cut-off value of normal baPWV was not available, high baPWV was arbitrarily defined as a value greater than the cut-off level between the third and fourth quartiles (> 75th percentile), which was 1437.5 cm/s for men and 1411.5 cm/s for women. All statistical analyses were performed using SPSS for Windows 23.0 (SPSS Inc., Chicago, IL, USA). P values less than 0.05 were considered statistically significant.

Table 1 shows the clinical and laboratory characteristics of the study population. Subjects were stratified into four groups based on their TyG index levels. Significant differences in metabolic parameters were observed among the groups. SBP, DBP, BMI, FPG, TC, TG, LDL-C, and HOMA-IR were all positively associated with the TyG index, while HDL-C was negatively associated with the TyG index. In addition, smoking status and the prevalence of diabetes were positively associated with TyG index quartiles.

We found that baPWV was positively associated with increasing TyG quartiles. Figure 1 shows the age-adjusted mean baPWV according to TyG index quartiles in the overall population, where Q1 = 1308.9, Q2 = 1338.4, Q3 = 1352.9, and Q4 = 1387.7 cm/s (P < 0.01). For men, the age adjusted mean baPWVs of the TyG index quartiles were 1344.9, 1358.2, 1360.7 and 1394.3 cm/s (P < 0.01), while for women the age-adjusted mean baPWVs in the four quartiles were 1291.8, 1318.0, 1337.9 and 1361.1 cm/s, respectively (P < 0.01).

Pearson’s correlation analysis was used to examine the relationship between the baPWV and clinical variables. The baPWV was significantly correlated with age SBP, DBP, BMI, TC, LDL-C, HDL-C, TG/HDL-C, LDL-C/HDL-C, non HDL-C HOMA-IR, and the TyG index (Table 2).

The association between the TyG index and high baPWV was further explored by categorizing the TyG index levels into quartiles and using the first quartile as a reference. Based on age-adjusted multivariable logistic regression analysis, with Q1 set as the reference, the TyG index levels for Q2, Q3, and Q4 were associated with increased ORs for high baPWV in all subjects. Both in men and in women, this relationship remained statistically significant after adjusting for confounding variables. In addition, higher HOMA-IR quartiles were significantly associated with increasing baPWV, and this relationship remained significant even after adjusting for metabolic variables. After adjusting for multiple risk factors, the OR (95% CI) for increased baPWV in the highest HOMA-IR quartile was 1.80 (1.17–2.78) in men and 1.46 (1.06–2.47) in women, whereas the OR of the TyG index in the highest quartile was 2.92 (1.92–4.44) in men and 1.84 (1.15–2.96) in women, respectively (Table 3).