Study design and population
The current data originated from the epidemiological data collected by the TIDE. This population-based, cross-sectional, epidemiological survey was performed in Jiangxi province from April to August in 2015. Sample size was estimated to meet generally recommended requirements for precision in a random sampling survey. Assuming the confidence interval was 95% and the MS prevalence 10%. The permissible relative sampling error was 4%. Therefore, the total sample size was 217. In this study, 2665 adults were invited and 2580 effectively attended, giving a response rate of 96.8%. The sample size achieved the minimum requirement for the sampling method.
A cluster sampling method was applied in this study (Figure 1). (1) According to the economical size of each city and county in Jiangxi province provided by Statistical Bureau, areas which fell in the 25th to 75th percentiles of the Gross Domestic Product (GDP) level were included. Besides, considering the data accessibility, sparsely inhabited districts with higher or lower economical sizes were excluded. (2) In the above areas, one urban city (Nanchang) and one rural county (Dexing) were chosen using a simple random sampling method. Afterwards, one community (Qingyunpu) and four villages (Diaozhong, Yincheng, Raoer and Huangbai) were picked using the above method. (3) The participants were restricted to local residents aged over 18 years old who had been living at the study sites for at least five years. Pregnant women were excluded. According to residential registration, all of the eligible candidates were included.
Questionnaire, anthropometric measurements and blood sample investigation
Firstly, individual demographic characteristics and disease history questionnaire were completed. Demographic and clinical information included age, sex, level of education, marital status, menstrual condition, family annual income, occupational status, smoking, history of diseases such as diabetes, hypertension, dyslipidemia and hyperuricemia, medications for diseases and family history of diabetes. In multiple logistic regression analysis, age was classified into 3 groups, including ≤40, 41-65 and ≥66 years old. Education was classified into 3 levels, including primary school and below, middle and high school, and college and above. Family annual income was graded into 4 levels: (1) < 30,000; (2) 30,000~50,000; (3) 50,000~100,000; (4) > 100,000 Chinese yuan (100 CNY= 14.56USD). Occupational status was categorized into student, worker (farmer), house worker (retired), clerk and others. Smoking was defined as at least 100 cigarettes consumed prior to the survey and was further divided into two groups on the basis of the frequency of cigarettes consumption (<1 cigarette per day or ≥1 cigarette per day). Amenorrhea was defined as not having a period for at least 6 months or 3 menstrual cycles.
Anthropometric data, including BP, heart rate (HR), WC, height and body weight, were measured by recommended standard procedures. In brief, BP and HR were average values of two separate measurements taken at 5-minute intervals. Weight and height were measured without shoes or heavy garments. BMI (kg/m²) was determined by dividing the weight (kg) by height (m) squared. WC was measured in the erect position at the middle of the lowest rib and the superior border of the iliac crest.
Blood samples were obtained after at least 10 hours of fasting to determine fasting blood chemistry parameters. Afterward, all subjects were given a standard 2 hour-75g oral glucose tolerance test (OGTT). All serum parameters were detected on a Mindray (Mindray Medical International Limited, China) automatic biochemistry analyzer. Serum total cholesterol (TC) and triglyceride (TG) levels were determined by enzymatic methods. Serum low density lipoprotein cholesterol (LDL-C) and high density lipoprotein cholesterol (HDL-C) amounts were measured by the direct method. Serum uric acid (UA) content was measured by the uric acid enzyme-peroxide enzyme coupling method. Fasting plasma glucose (FPG) and OGTT-2h plasma glucose (2h PG) were determined by the glucose oxidase method. Glycosylated hemoglobin A1c (HbA1c) was measured by high pressure liquid chromatography. All the procedures were executed by experienced laboratory technicians.
Diagnosis of metabolic syndrome
MS was diagnosed based on IDF or CDS criteria (Chinese specific)[9, 14]. These two criteria are described in Supplementary Table 1.
Data collection and Statistical Analyses
In this study, data collection was performed by trained medical professionals at the Second Affiliated Hospital of Nanchang University. Data analysis was limited to individuals who had completed all procedures, comprising 1322 (51.2%) males and 1258 (48.8%) females; 1360 (52.7%) participants lived in urban areas, and 1220 (47.3%) in rural regions.
An EpiData (EpiData Association, Odese, Denmark) database was established, and all data were analyzed by SPSS (Statistical Program for Social Sciences, version 20.0). Continuous variables were described as mean (standard error, SE) and analyzed by student’s t-test. Categorical variables are presented as number and percentage, and analyzed by the Chi-square test. The official 2010 census data of China was used to determine age-standardized ratios.
We analysed the associated factors for MS with logistic regression models. We reported the odd ration (OR) with 95% CI. Covariates included in the multivariable logistic regression models were profession (student [reference] vs worker vs clerk vs houseworker vs others), family annual income (below 30,000 [reference] vs 30,000-50,000 vs 50,000-100,000 vs above 100,000), age (18-40 [reference] vs 41-65 vs above 66), education level (primary school or below [reference] vs middle or high school vs college or above), family history of diabetes (yes [reference] vs no),smoke (no or less than 1 cigarette/d [reference] vs more than 1 cigarette/d), Menses condition (menopausal state [reference] vs un-menopausal state). We also investigated whether the associated factors for MS varied by region and gender by stratified multivariable analyses (in urban males, urban females, rural males and rural females respectively).