Prevalence of chronic kidney disease markers: Evidence from a three-million married population with fertility desire in rural China

We aimed to assess the prevalence of chronic kidney diseases (CKD) markers among the married residents with fertility desire in rural China. Demographic and clinical data were collected from the National Free Pre-Conception Health Examination Project. Estimated glomerular filtration rate (eGFR) < 60 mL/min//1.73 m2, proteinuria, and hematuria were defined as markers of CKD. GFR was evaluated by using serum creatinine level and the Asian-modified CKD epidemiology collaboration equation. Automated urine dry chemical and microscopic analyses were employed to identify proteinuria and hematuria. The prevalence of CKD markers was 2.92% in the 3,091,379 participants. eGFR < 60 mL/min//1.73 m2, hematuria and proteinuria was observed in 0.85%, 1.41% and 0.71%, respectively. The prevalence of CKD markers varied greatly across different geographical locations, which was the highest in the Eastern Region (3.86%; 95% confidence interval [CI]: 3.81–3.91%), moderate in the Central Region (2.80%; 95% CI: 2.77–2.82%), and lowest in the Western Region (2.62%; 95% CI: 2.59–2.65%). Hypertension, obesity, positive hepatitis B virus surface antigen (HBsAg), age (increased by every 5 years), female gender, and living area were potential risk factors for CKD. In rural China, the prevalence of CKD markers in the married couples with fertility desire is low.


Screening Protocol and Evaluation Criteria. According to the Kidney Disease Outcomes Quality
Initiative (KDOQI) clinical practice guideline, estimated glomerular filtration rate (eGFR) < 60 mL/min//1.73 m 2 , proteinuria, and hematuria were defined as markers of CKD 30 . As our study did not repeat the measurements until three months later, the chronicity criterion of CKD was not applied to this definition. The morning spot urine samples were collected to identify hematuria and proteinuria by means of automated urine dry chemical analysis. Samples with positive hematuria for one or more times were reexamined through microscopic analysis within 2 hours by qualified technicians in the county clinical laboratories. Presence of 3 or more red blood cells per high-power field was considered abnormal. Subjects with hematuria and pyuria were considered to have urinary tract infection and excluded from data analysis. Women undergoing menstruation were also excluded from urine analysis.
Blood samples were collected by means of venipuncture after an overnight fast of at least 10 hours and immediately sent to the county clinical laboratories. Fasting blood glucose (only for women), serum creatinine (Scr), and hepatitis B virus surface antigen (HBsAg) were examined by qualified technicians. The reagent kits approved by the China Food and Drug Administration were chosen by the local laboratories on their preference, and were double checked by National Center of Clinical Laboratories for Quality Inspection (NCCLQI). Sensitivity, specificity, and κ value of the selected reagents from all the involved county laboratories were higher than 95%. Provincial Center of Clinical Laboratories for Quality Inspection carried out casual inspections and NCCLQTD conducted an external quality assessment (EQA) for quality control 31 .
Serum creatinine was measured by enzymic method with Enzymatic Creatinine-2 Reagents (Siemens Healthcare Diagnostics Inc.) and isotope dilution mass spectrometry (IDMS) method as the reference standard. Hepatitis B virus surface antigen (HBsAg) was detected by Siemens ADVIA 2400 Chemistry System with reagents produced by Abbott (Abbott Park, IL, USA) as the reference standard. eGFR was calculated by using Chinese modification of diet in renal disease (C-MDRD) equation 32 and Asian-modified chronic kidney disease epidemiology collaboration (CKD-EPI) equation 33 , respectively as follows: We adopted Asian-modified CKD-EPI equation to evaluate CKD prevalence. Arterial blood pressure was measured in sitting position by sphygmomanometer three times at 5 minute intervals after the subjects had rested for at least 15 minutes. The mean of the three readings were calculated unless the difference between readings was greater than 10 mmHg, in which case the mean of the two closest readings was applied. Hypertension was defined as systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg, or by any self-reported history of hypertension. Because fasting blood glucose and postprandial blood glucose of the men were not measured, for men, "diabetes" was defined as history of diabetes. For women, diabetes was defined as history of diabetes or fasting blood glucose ≥7.0 mmol/L. The body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared. According to Chinese criteria of weight for adults 34 , 18.5 kg/m 2 ≤ BMI < 24.0 kg/m 2 was in normal range, BMI < 18.5 kg/m 2 represented underweight, 24.0 kg/m 2 ≤ BMI < 28.0 kg/m 2 represented overweight. Obesity was defined as BM I ≥ 28.0 kg/m 2 .
Statistical analysis. Continuous variables were analyzed by one-way ANOVA. Categorical variables were analyzed by χ 2 test. The prevalence of CKD markers, eGFR < 60 mL/min//1.73 m 2 , hematuria and proteinuria were analyzed with age groups by spearman correlation test. To investigate the risk factors and indicators for CKD, logistic regression models were applied. The crude and multivariable adjusted odds ratios (ORs) were reported. Covariates included in the multivariable logistic regression models were hypertension (no vs. yes), obesity (no vs. yes), HBsAg positive (no vs. yes), gender (female vs. male), and region (eastern region: 0; central region: 1; and western region: 2). Age was defined as categorical variable with a 5-year interval, and 20-24 years was treated as a dummy variable in the Logistic Regression Model. All statistical data were analyzed after removing the missing items. The significant level was 0.05 and all of the analyses were two-sided tests. Statistical analyses were performed with SPSS version 21.0.

Results
Demographic and clinical characteristics of the studied population. 3,091,379 participants filled in the questionnaire. Rate of loss of participants due to uncompleted blood test and urinalysis was 3.58%, 0.51% and 3.52%, respectively, in the Eastern Region, Central Region and Western Region. The average age was 27.04 ± 7.10 (standard deviation [SD]) years. As shown in Table 1, the proportion of female was 51.70%. The surveyed population mainly engaged in agriculture. Nearly 2/3 of people had junior middle-school education. The average level of Scr was 77.06 ± 20.35 µmol/L. 4.06% of the population was suffering from hypertension, with males more susceptible than females. More females had a history of diabetes than males (0.0214% vs. 0.0156%). The prevalence of diabetes was 1.41% for females. 5.64% of the population had positive HBsAg, with males more susceptible than females. The prevalence of underweight in females was 3 times higher than that in males, while the prevalence of overweight and obesity in males was twice as that in females.
The prevalence of CKD markers by Asian-modified CKD-EPI equation for eGFR was 2.92% in the surveyed population (1.81% and 3.95%, respectively, in males and females). If hematuria was excluded, the prevalence of CKD markers was dropped to 1.54% (1.08% and 1.96%, respectively, in males and females). By adopting Asian-modified CKD-EPI equation for eGFR, the prevalence of eGFR below 60 mL/min//1.73 m 2 was 0.85% (0.43% and 1.24% in males and females, respectively). By adopting C-MDRD formula, the prevalence of eGFR below 60 mL/min//1.73 m 2 was 1.50% (0.98% and 1.91% in males and females, respectively). The prevalence of hematuria was 1.41% (2.03% vs. 0.74% in females vs. males). The prevalence of proteinuria was 0.71% (0.74% vs. 0.66% in females vs. males). Table 2, the prevalence of CKD 1-5 stage based on the estimated GFR by Asian-modified CKD-EPI equation were 1.66%, 0.40%, 0.79%, 0.03% and 0.03% respectively. There were differences between the prevalence of the males and the females in every CKD stage. Table 3, the prevalence of CKD markers varied greatly due to different geographical locations, which was the highest in the Eastern Region (3.86%), moderate in the Central Region (2.80%) and the lowest in the Western Region (2.62%). The prevalence of hematuria and proteinuria in Eastern Region was significantly higher than that in the Central and Western regions. Hypertension, obesity and positive HBsAg were most frequently observed in the Eastern Region.

The prevalence of CKD markers in different geographical locations. As shown in
The prevalence of CKD markers in different age groups. As shown in Fig. 1, the prevalence of CKD markers except proteinuria increased with age. The prevalence of CKD markers in the aged 46-49 group was two times higher than that in the aged 20-25 group. The prevalence of eGFR below 60 mL/min//1.73 m 2 (by Asian-modified CKD-EPI equation) in the aged 46-49 group was four times higher than that in the aged 20-25 group.
Scientific RepoRts | 7: 2710 | DOI:10.1038/s41598-017-02355-2 The prevalence of CKD markers in the population with chronic diseases. As shown in Table 4, the prevalence of CKD markers was significantly higher in subjects with hypertension, diabetes, positive HBsAg, and obesity compared to their counterparts. Table 5 listed the crude and adjusted odds ratios (ORs) for CKD as well as potential indicators. Hypertension, obesity, positive HBsAg, female and age (increased by every 5 years) were independent risk factors for eGFR < 60 mL/min//1.73 m 2 . Hypertension, obesity, positive HBsAg, female, age (increased by every 5 years) and region were independent risk factors for hematuria. Independent risk factors for proteinuria included hypertension, obesity, positive HBsAg, female and region. Hypertension, obesity, positive HBsAg, female, age (increased by every 5 years) and region were independent risk factors for CKD.   CKD. Second, the prevalence of risk factors for CKD diverges greatly in different investigations. Participants in Mainland China study were characterized by an incidence of hypertension as 35.4% and self-reported diabetes as 7.4% 22 . The prevalence of hypertension and diabetes reported by the study in Taiwan was 20.7% and 5.4%, respectively 5 . Those values outnumber the corresponding ones in our study. A sampling bias stemming from more subjects with chronic diseases may result in higher prevalence of CKD in some previous studies. Third, different methods were applied for eGFR assessment in different studies. The current study adopted Asian-modified CKD-EPI equation instead of C-MDRD equation to evaluate GFR. C-MDRD formula had been used in previous investigations in China, which was modified on the basis of CKD population in patients with eGFR < 60 mL/ min//1.73 m 2 and in the elderly 3,31 . But C-MDRD formula tends to underestimate normal kidney function and overrate the prevalence of CKD 35 . Asian-modified CKD-EPI equation was developed from subjects regardless of kidney diseases and it was more precise and accurate than C-MDRD equation, especially at a higher eGFR [36][37][38][39] . For example, in Romania the prevalence of CKD was 0.95% by C-MDRD equation whereas 0.64% by CKD-EPI equation in those aged between 18 and 44 3 . In this study, the ratio of eGFR below 60 mL/min//1.73 m 2 evaluated by C-MDRD equation was twice as that by Asian-modified CKD-EPI equation. A systemic underestimation of GFR by the C-MDRD equation may account for a higher prevalence of CKD in the previous studies in China. Forth, none of the previous studies used hematuria in defining CKD according to KDIGO guideline 5,22 . If hematuria was excluded from the definition, the prevalence of CKD markers was decreased to 1.54% in married population of reproductive age in rural China. 40 . Several studies have revealed a strong association of CKD with increased morbidity and mortality of hepatitis B 18 . At present, the studies on risk factors for CKD in China seldom involve HBV infection. We found that the prevalence of CKD markers is higher in the participants with positive HBsAg than those with negative HBsAg; and importantly, positive HBsAg is an independent risk factor for CKD. The result informs that universal hepatitis B vaccination program for infants enacted in 1992 to control HBV infection might also be beneficial in controlling CKD in China. According to the economic development level and geographical locations, Mainland China has been divided into Eastern Region, Central Region and Western Region. There have been major differences in these three regions. The Eastern Region is coastal, where local economy is developed. In the Western Region, geography is complex, natural disasters happen frequently and development lags behind. The Central Region lies inland with moderate economy development level. The Gross Domestic Product (GDP) per capita in the Eastern Region is 1.8 and 1.7 times higher than that of the Western Region and the Central Region, respectively, in 2012 41 . Different from other countries and regions where the prevalence of CKD increased with low SES and poverty 5,8,9,27 , our study finds that the prevalence of CKD markers in the Eastern Region exceeds that of both the Central and the Western regions in China. Another study also found that economic development was positively associated with albuminuria in rural China 22 . There are several explanations for different prevalence of CKD markers among different geographical locations. First, rapid economic development and urbanization result in changes of life style in the Eastern rural region. The highest prevalence of hypertension and obesity stemming from modern life style may explain the highest prevalence of CKD markers in the Eastern rural region. Second, hepatitis B infection as an independent risk factor for CKD is more prevalent in the Eastern Region than that in the Central or Western regions. Third, heredity diversity may be another important factor.

Hepatitis B virus (HBV) infection is highly endemic in rural China
Regarding gender distribution, inconsistent results have been reported. In the studies of Guangzhou, Shanghai and Mainland China, prevalence of CKD in women was all higher than that in men 22  predominance had been observed in other countries 3,8,42 . The authors explained that the discrepancy was attributed to naturally lower eGFR in women. But in Japan, males were more susceptible to CKD due to a higher uric acid level 4 . In Taiwan, the prevalence of CKD in men was also higher than that in women 5 . In our study, the incidence of hypertension, obesity, and HBV infection, as well as the mean age of females were all lower than those of males, but the prevalence of CKD markers in females was as 2 times as that in males. Even after hematuria was excluded from the definition of CKD, the prevalence CKD markers in females remained much higher than that in males. Different congenital susceptibility to kidney damage may contribute to different prevalence of CKD markers between males and females. This study has some limitations. First of all, according to KDIGO guideline, CKD should have indicators of kidney damage over three months. In this research, all the indicators of CKD were tested from single measurement, because it was extremely difficult to conduct repeated examination in such a large-scale investigation. Lack of confirmatory dipstick urinalysis and demonstrated chronicity of eGFR below 60 mL/min//1.73 m 2 may lead to remarkedly overrated CKD prevalence 42 . Although some authors would believe that most of the previous studies using single measurement for serum creatinine and dipstick urinalysis might have been equally affected 43 , this limitation would result in variation in prevalence of CKD across studies. The actual prevalence of CKD in married population with fertility desire in rural China would be even lower than that in our study. Second, applying dipstick urinalysis as screening experiment might have missed diagnosis of kidney damage 20,44 , thus underestimating the prevalence of CKD. However, in the nationwide investigation, dipstick urinalysis has been used as the most convenient, feasible and cost-effective tool to identify kidney damage, especially in rural areas. Dipstick urinalysis is sensitive enough to detect urinary protein equivalent to microalbuminuria 45 . Glomerulnephtitis remains the most common chronic kidney disease and the leading cause of end-stage renal disease in China 46 . So dipstick urinalysis, rather than urinary albumin to creatinine ratio, was selected to diagnose hematuria and proteinuria 47 . The higher prevalence of hematuria among young women might be from renal or urological origin, but not necessarily caused by CKD, which requires further investigation. Third, there was possible potential selection bias that would underestimate the prevalence of CKD. The population with serious CKD might not enroll in the survey because of having no fertility desire. Unmarried people in rural China were not included, which might be another major limitation in the current research. Forth, fasting blood glucose was measured only for women, which was another major limitation. The association of diabetes with CKD prevalence stratified by gender requires further investigation.
Despite the limitations, with random sampling and a 3-million sample size, this study had a high power that the power of CKD markers (by Asian modified CKD-EPI equation) and CKD markers (by Asian modified CKD-EPI equation and excluding hematuria) between males and female was 100%. For the first time, our study has reported the prevalence of CKD markers in married population with the mean age of 27 years in rural China. Our study also highlights variation in the prevalence of CKD across regions and potential factors attributed to this variation. In rural China, the prevalence of CKD markers in married population with fertility desire is low. Different from other studies in which the prevalence of CKD was higher in people with low socioeconomic status, a higher prevalence of CKD markers is observed in the developed Eastern Region. Hypertension, obesity, positive HBsAg, female, age and living area were independent risk factors for CKD. These results indicate that economic development and changing life style have affected the epidemiology of CKD. Better understanding and management of risk factors for CKD is critical to prevent this disease.