Study design and participants
Beijing health management cohort study (BHMC) is an ongoing community-based prospective cohort study established in 2008, specially designed to investigate the risk factors and blood biomarkers for metabolism-related diseases, such as hypertension and diabetes. The BHMC study was conducted in Beijing, China, and the recruited participants underwent a comprehensive annual health examination, face-to-face questionnaire survey and blood sample collection. Details of the study design have been described previously [20]. Of 62,311 participants who underwent health examination in 2013 or 2014 (at baseline), 8,063 were diagnosed with hypertension. To minimize the possible effect of reverse causality, we excluded 2,167 participants with an ankle-brachial index (ABI) <0.90 and 2,984 participants with brachia-ankle pulse wave velocity (baPWV) <1800 cm/s, Then, 688 participants using glucose-lowering medication, 231 participants using lipid-lowering medication, 784 participants unable to collect the required information at baseline and 11 participants lost to follow up were further excluded. Finally, this study was restricted to a subset of 1201 participants with complete data and considered in the final analyses as shown in Figure 1.
This study was in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of Capital Medical University. All participants provided their written informed consents before taking part in this study.
Data collection and definitions
Sociodemographic characteristics and lifestyles were collected via a standard questionnaire by trained staffs, including age, sex, smoking status, drinking status, physical activity levels and previous medical diagnoses. Smoking and drinking status was defined as ‘current’ and ‘never or former’. Physical activity was classified as ‘>80 minutes per weak’ and ‘<80 minutes per week or none’. The physical and biochemical examination data were acquired from the electronic medical record system. Body mass index (BMI) was calculated weight (in kilograms)/height*height (in meters squared). Waist hip ratio (WHR) was defined as waist (in centimeter)/hip (in centimeter). Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were presented as the average of two measurements on the right arm using a sphygmomanometer after resting for at least 10 min. Hypertension status was defined as SBP ≥140 mmHg or DBP ≥90 mmHg or use of any anti-hypertension medication, or self-reported history of hypertension diagnosis.
Blood samples were stored and measured in the central laboratory of Beijing Xiaotangshan Hospital using the Olympus Automatic Biochemical Analyzer (Hitachi 747; Tokyo, Japan). Serum total cholesterol (TC), triglyceride (TG), high density lipoprotein cholesterol (HDL-C), and low density lipoprotein cholesterol (LDL-C) were measured with the enzymatic color-metric method. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI 2009) serum creatinine equation [21]. Fatty liver was diagnosed by ultrasonic examination. The fasting blood glucose (FBG) was defined as the glucose concentrations before breakfast after overnight fasting (no food, except drinking water, for at least 8-10 hours), while two-hour postprandial blood glucose (PBG) was measured after 2 hours from the beginning of fixed meals through finger blood test. Diabetes was defined as FBG ≥7.0 mmol/l or PBG ≥11.1 mmol/l or the glycated hemoglobin (HbA1c) ≥6.5% or use of any glucose-lowering medication or self-reported history of diabetes.
The TyG index was denoted as ln[TG (mm/L)*fasting glucose (mm/L)/2]. The TG/HDL-C ratio was calculated as TG (mm/L) divided by HDL-C (mm/L).
Assessment of arterial stiffness
Arterial stiffness refers to the measurement of baPWV, which is a simple, noninvasive, automatic method and widely used in the clinical practice and large population-based studies. The baPWV was measured with Omron Colin BP-203RPE III device (Omron Health Care, Kyoto, Japan). After more than 5 minutes’ rest in supine position, 4 cuffs were wrapped around bilateral brachia and ankles, then connected to a plethysmographic sensor and oscillometric pressure sensor. ABI refers to the ratio of the ankle SBP divided by the brachial SBP. Semiconductor pressure sensors were used to assess the transmission time between the initial rises in both the brachial and tibial arteries waves in order to record pressure waveform. The distance between sampling points of baPWV was determined based on the height of the subjects. The baPWV was calculated according to the formula (La-Lb)/Tba. La is the path length from the heart to the ankle, Lb is the path legth from the heart to the brachium, and Tba is the time interval between the brachial and ankle waveform. The measurement was performed twice by two trained technicians, and the average value of the left and right sides was calculated as the final result. The incidence of arterial stiffness was confirmed with a BaPWV >1800 cm/s [22].
Statistical analysis
Baseline characteristics are presented as the mean (standard deviation, SD), median [interquartile range, IQR] or number (percentage). Differences between incident arterial stiffness and non- arterial stiffness groups were compared using Student’s t-test or Mann-Whitney U test for continuous variables and Pearson’s chi-sqaure test or Fisher’s exact test for categorical variables.
The Cox proportional hazards regression models were used to estimate the association of TyG index and TG/HDL-C ratio with arterial stiffness in hypertensive population. TyG index and TG/HDL-C ratio categorized in quartile were also analyzed. The restricted cubic spline function was performed to illustrate to linear or non-linear correlation. To adjust for potential confounding factors, three models were established as follows: Model 1 adjusted for age and sex; Model 2 adjusted for age, sex, BMI, WHR, SBP and DBP; Model 3 further adjusted for eGFR, uric acid, homocysteine, smoking status, drinking status, physical activity, diabetes, fatty liver and use of anti-hypertension medication. The hazard ratio (HR) and 95% confidence interval (CI) were presented. The partial Spearman’s correlation coefficients between TyG index and TG/HDL-C ratio, and other common cardiometabolic risk factors were presented, after adjusted for age and sex.
To identify the effect modifying factors, we conduct subgroup analyses. The factors, by which the observed association was significantly amplified or concealed, were referred as potential effect modifiers. All the analyses presented above were conducted using R software (version 3.6.3). The difference was considered statistically significant at two-side P <0.05.
Data availability
The data of this cohort study are available to researchers on request by contacting with the corresponding author (Dr. Guo).