Study subjects
From December 2017 to June 2018, the subjects were recruited from 3 communities in 3 districts of Zhengzhou: the Erqi, Zhongyuan and Jinshui districts. There are 12 administrative districts with a total population of approximately 10 million in Zhengzhou (data available on http://tjj.zhengzhou.gov.cn/). A multistage, stratified cluster sampling method was employed to select participants over 18 years old from the general population. In the first stage, 3 districts were randomly selected from 6 urban districts. In the second stage, one representative community in each district was selected according to the proportion of permanent residents. In the final stage, all permanent residents who satisfied the inclusion criteria and agreed to sign the informed consent were recruited in this study. Altogether, a total of 6000 subjects aged 18 years or older were selected from 3 communities, and 5231 subjects completed the survey and examination, corresponding to a response rate of 87.2%. Two hundred and thirty three subjects were dropped because they didn’t respond and 536 subjects were dropped because they didn’t complete the questionnaire.
Measurements and definitions
Data were collected by face-to-face interviews in examination centers at community health stations. All subjects completed a questionnaire that collected information about their sociodemographic status (e.g., age, gender, ethnicity, education, etc.), personal and family health history (e.g., diabetes, hypertension, hepatitis, etc.), lifestyle behaviors (e.g., smoking, alcohol consumption, physical activity, etc.) and awareness and control of chronic noncommunicable disease (e.g., diabetes, hypertension, dyslipidemia, etc.) with assistance of trained practitioners, doctors and nurses. Anthropometric measurements, such as height, weight and blood pressure (BP), were obtained. Height and weight were measured while the participants were in light clothing without shoes, and body mass index (BMI, kg/m2) was subsequently calculated. According to the Chinese “Criteria of weight for adults (No. WS/T 428-2013)” (available on http://www.nhfpc.gov.cn), BMI was divided into four levels: underweight (< 18.5), healthy weight (18.5 – 23.9), overweight (24 – 27.9) and obesity (≥ 28). BP was measured using an electronic sphygmomanometer (Omron HEM-7071A, Japan) three times in one-minute intervals. The mean value of the three BP readings was used for statistical analysis unless the difference between the readings was higher than 10 mm Hg, in which case the mean value of the other two closest results was calculated. In addition to participants’ self-reported use of antihypertension medications in the past 2 weeks, hypertension was defined as participants with an average systolic BP (SBP) ≥ 140 mm Hg and/or an average diastolic BP (DBP) ≥ 90 mm Hg [19]. Subjects with hypertension were considered to have controlled BP if SBP < 140 mm Hg and DBP < 90 mm Hg.
After at least 8 hours of overnight fasting, venous blood specimens were collected in vacuum tubes without an anticoagulant. Serum concentrations of creatinine, uric acid, total cholesterol, triglycerides, high-density lipoprotein and low-density lipoprotein were measured using enzymatic colorimetry on a Cobas C 701 (Roche). The fasting plasma glucose (FPG) level was estimated by the glucose oxidative method (GOD-PAP). Urinary albumin and creatinine were measured from a fresh morning spot urine sample. Albuminuria was measured with an immune-turbidimetric test. Urinary creatinine was evaluated by Jaffe’s kinetic method. The urinary albumin to creatinine ratio (ACR, mg/g) was calculated automatically.
The estimated glomerular filtration rate (eGFR) was calculated by serum creatinine using the 2009 CKD-EPI creatinine equation [20]. According to the 2012 KDIGO classification, eGFR was classified into 5 stages, in which stage 3 was further divided into stages 3a and 3b with an eGFR of 45 mL/min/1.73 m2 as the cut-off value. ACR was classified as follows: A1, < 30 mg/g; A2, 30 – 300 mg/g; and A3, > 300 mg/g [21]. Albuminuria was defined as an ACR ≥ 30 mg/g. Indicators of renal damage were the presence of an eGFR less than 60 mL/min/1.73 m2 or albuminuria. CKD was defined as the presence of one or two indicators of renal damage.
Diabetes was defined according to the 2009 American Diabetes Association (ADA) guidelines: 1. FPG ≥ 7.0 mmol/L; 2. self-reported use of insulin or antidiabetic medications in the past 2 weeks; and 3. self-reported previous diagnosis of diabetes by a physician. Diabetic subjects with systolic blood pressure >140 mmHg, albuminuria or an eGFR less than 60 mL/min/1.73 m2 were classified as having DKD. Control of diabetes was defined as FPG maintained at less than 7.0 mmol/L for the past 7 days. Dyslipidemia was considered the use of antidyslipidemia medications during the last 2 weeks or the presence of one or more abnormal serum lipid concentrations according to the Chinese guidelines for the prevention and treatment of dyslipidemia in adults: 1. Total cholesterol > 6.22 mmol/L; 2. Triglycerides > 2.26 mmol/L; 3. High-density lipoprotein cholesterol < 1.04 mmol/L; 4. Low-density lipoprotein cholesterol > 4.14 mmol/L[22]. Hyperuricemia was defined as a plasma uric acid concentration > 422 µmol/L for men and > 363 µmol/L for women.
Education was divided into 3 levels: 1. Primary school or lower; 2. junior middle school; and 3. senior high school or above. Diets rich in fruits and vegetables were considered diets with a daily average consumption of more than 500 g of fruits and vegetables. A high-fat diet was defined as a diet with a daily average consumption of livestock and poultry of more than 75 g [23]. Physical activity was classified as low, moderate or high according to the international physical activity questionnaire (IPAQ 2001) [24].
Statistical analysis
Epidata software (version 3.1) was used for data entry and management. All statistical analyses were performed with SAS 9.1 (SAS Institute, Cary, NC, USA) and GraphPad Prism 6 (GraphPad Software, Inc., La Jolla, CA, USA) for Windows. A p-value < 0.05 was considered statistically significant. Data are expressed as the mean ± SD, median with range or frequency with percentage, as appropriate. Intergroup comparisons were performed using the Pearson chi-square test for categorical variables and Student’s t-test, the Mann-Whitney U-test or the Wilcoxon test for continuous variables, as appropriate. The standard population of this study was based on data from the China Population Sampling Census in 2009 (data available on http://www.stats.gov.cn/).
The crude and adjusted prevalences of reduced eGFR (eGFR < 60 mL/min/1.73 m2), albuminuria, DKD and CKD were reported. Both binary and ordinal logistic regressions were employed to explore the associations between indicators of renal damage and the relevant covariates. In binary logistic regression, crude and multivariable adjusted odds ratios (ORs) with 95% confidential intervals (CIs) were calculated. The covariates involved in our multivariable logistic regression model were age (in 10-year intervals), gender (men versus women), education [≤ primary school (reference) versus junior high school versus ≥ senior high school], current smoker (yes versus no), alcohol consumption (yes versus no), BMI [healthy weight (reference) versus underweight versus overweight versus obesity], diabetes (yes versus no), hypertension (yes versus no), dyslipidemia (yes versus no), and hyperuricemia (yes versus no). According to the prognosis of CKD by GFR and albuminuria category from the 2012 Kidney Disease: Improving Global Outcomes (KDIGO) guidelines, two models were used to calculate the data in the ordinal logistic regression. In model 1, we analyzed data from subjects with an eGFR > 60 mL/min/1.73 m2 and different levels of ACR (A1 – A3, n=5099) [21]. In model 2, data from all participants were divided into 4 groups as follows: low risk (no CKD); moderately increased risk; high risk; and very high risk [21]. The results of tests of parallel lines indicated that the two models were statistically executable (both P values > 0.05).