The data were obtained from the Korean National Health and Nutrition Examination Survey (KNHANES) conducted by the Korean Centers for Disease Control and Prevention. The KNHANES is a nationwide cross-sectional epidemiologic survey with a probability-cluster, multistage, and stratified sampling design. A total of 41,347 individuals participated in the 2007–2011 KNHANES. Of these, individuals who met the following criteria were excluded: (1) those under 30 years of age, (2) those lacking information about the administrative divisions needed to identify their air pollution exposure, (3) those without the phenotypic information needed for the eGFR calculation, such as age, sex, and serum creatinine level, and (4) those who reside in Jeju Island, which is an island region that is environmentally different from land area. Finally, a total of 24,407 adults were included in the statistical analysis. Written consent was obtained from all individuals before participating in the survey and their data were anonymized. This study was approved by the institutional review board of the Seoul National University Hospital Biomedical Research Institute.
Renal function measurement
Blood samples were drawn from survey participants after fasting for at least 8 h. The serum creatinine level of each sample was measured by a professional blood testing agency. The individual’s eGFR, a representative value indicating renal function, was calculated using the Modification of Diet in Renal Disease (MDRD)-4 equation: GFR (mL/min per 1.73 m2) = 175 x SCr-1.154 x age-0.203 × 1.212 (if the individual was black) × 0.742 (if female), where SCr is the serum creatinine level. The CKD was defined as eGFR < 60 mL/min/1.73 m2, which represents a reduction in renal function of half or more of the normal level .
Air pollution exposure
To estimate each individual’s exposure to ambient air pollution, we used monitoring data for 24-h concentrations of ambient air pollution collected from January 1, 2007, to December 31, 2011, by the Ministry of the Environment of Korea (https://www.airkorea.or.kr). These data were obtained from about 300 atmospheric monitoring sites nationwide in South Korea. The ambient air pollutants analyzed in the present study were PM10, NO2, SO2, and CO. The KNHANES survey data do not provide the actual home addresses of the survey participants, required to estimate the concentrations of air pollution, to which they were exposed. We therefore calculated the annual average concentrations of air pollutants for the 16 administrative divisions of South Korea (7 metropolitan cities and 9 provinces). Of these, one province (Jeju Island), which differs from the other administrative divisions environmentally and culturally, was excluded from this study. The individuals’ residential division codes were then used to link them to the annual average pollutant levels in 15 administrative divisions.
We investigated potential covariates for the associations between ambient air pollution and renal function from the KNHANES survey. Demographic data, including age, sex, household income, education level, smoking status, and alcohol consumption, were obtained via a questionnaire. Smoking status was coded as three categories: never-smoker, former-smoker, and current-smoker. Alcohol consumption was assessed according to the frequency of drinking alcohol per month over the previous year and was classified into four categories: never, less than once a month, twice or three times a month, and more than 4 times a month. We divided the daily protein intake (g) into four levels using quartiles, and the fourth quartile group (protein intake > 81.7 g) was defined as high protein intake group. Residential regions were classified into two categories: urban and rural. Anthropometric data such as height and weight were also obtained, and body mass index (BMI) was calculated as the weight in kilograms divided by the square of the height in meters (kg/m2). Clinical data for this study, including total cholesterol and fasting glucose, were obtained from blood. Diabetes was defined as fasting glucose ≥126 mg/dL or taking diabetes treatment medication or taking insulin by injection, or physician diagnose. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured three times, and the mean values of the second and third measurements were used in the analysis. Hypertension was defined as SBP ≥ 140 or DBP ≥ 90 mmHg or an individual taking hypertensive drugs for more than 20 days a month.
We used Pearson’s correlation analysis to assess the correlations between the ambient air pollutants. Independent sample t-tests were conducted to evaluate the difference in annual mean air pollution levels between the two types of residential areas (urban and rural), as well as to compare the annual mean air pollution levels between CKD and normal groups. Multiple linear regression analysis was performed to identify associations between ambient air pollution and eGFR level, with the results indicated as beta coefficients and 95% confidence intervals (CIs) for renal function. The associations between ambient air pollution and CKD were determined by multiple logistic regression analyses, estimating the odds ratios (ORs) and 95% CIs of each air pollutant for CKD. These statistical estimates were scaled to the interquartile range (IQR) for each pollutant (10 μg/m3 for PM10, 12 ppb for NO2, 1 ppb for SO2, and 0.1 ppm for CO). We estimated three statistical models using gradually adjusted methods: unadjusted model (without covariates); Model 1, adjusted for age, sex, household income quartile, education level, smoking, alcohol consumption, high protein intake, survey year, and residential region; and Model 2, adjusted for Model 1 plus BMI, total cholesterol, fasting glucose, diabetes, systolic blood pressure, and hypertension. All analyses were performed with SAS 9.3 (SAS Institute, Cary, NC, USA).