Background We aimed to examine the risk factors for renal function decline (RFD) in a community-based cohort of a rural Chinese population with normal kidney function (estimated glomerular filtration rate, eGFR ≥60 mL/min/1.73 m2), both for the population as a whole and stratified by sex.
Methods 2518 participants were included in the current analysis. RFD was defined as follows: a drop in the eGFR category accompanied by a 25% or greater drop in eGFR from baseline; or a sustained decline in eGFR of more than 5 mL/min/1.73 m2/year.
Results The incidence rate of RFD was 8.7% (women 7.4% and men 9.8%). In the multivariable logistic regression model, the ORs (95% CI) of developing RFD was 1.60 (1.01 to 2.54) for men versus women, and 1.51 (1.09 to 2.08) for participants with obesity or abdominal obesity versus none (1.35 (0.85 to 2.14) for men, and 1.65 (1.04 to 2.64) for women). However, prehypertension (OR=1.64; 95% CI 1.02 to 2.63) or hypertension (2.05; 1.21 to 3.47), higher mean blood pressure (≥90 vs <80 mm Hg, 2.63; 1.11 to 6.20), higher pulse pressure (≥50 vs <40 mm Hg, 2.00; 1.26 to 3.18), lower high-density lipoprotein cholesterol (<0.9 vs ≥0.9 mmol/L, 2.65; 1.08 to 6.50) and low physical activity levels (vs high, 3.11; 1.59 to 6.10) were major risk factors for RFD in men. Current smoking (3.22; 1.22 to 2.64) and worse self-reported health (vs better, 2.57; 1.20 to 5.50) were major risk factors for RFD in women.
Conclusions Our findings suggested that sex-specific risk factors should be considered in prevention of RFD in the Chinese rural population with normal kidney function.
- Cohort studies
Statistics from Altmetric.com
Chronic kidney disease (CKD) is now recognised as a worldwide public health problem that substantially elevates the risk of cardiovascular events as well as end-stage renal disease (ESRD) and other complications. The prevalence of CKD is high and is increasing in China. A recent national cross-sectional survey in the Chinese adult population showed that the overall prevalence of CKD was 10.8% and that the prevalence was higher in rural areas (11.3%) than in urban areas (8.9%).1 Furthermore, even a mild decline of renal function is associated with an increased risk for ESRD and mortality.2 ,3
These statistics highlight the importance of identifying and understanding the related risk factors of renal function decline (RFD) as a basis for designing treatment strategies for preventing the development and progression of CKD. Most importantly, there is increasing but inconclusive evidence of sex-specific differences in the association of some risk factors, such as obesity, smoking and blood pressure, with the risk of RFD,4 incident CKD5 or the change in estimated glomerular filtration rate (eGFR).6 To further address these issues, we examined the risk factors, especially modifiable ones, for RFD in a community-based cohort of the rural Chinese population without known cardiovascular disease (CVD), diabetes, hypertension, cancer and with normal renal function (eGFR ≥60 mL/min/1.73 m2), both for the population as a whole and stratified by sex.
Participants and methods
Study population and data collection
Study participants were from an epidemiological study of metabolic syndrome conducted during 2003–2005 in rural communities (Dongzhi and Wangjiang) in Anqing, Anhui province of China. A detailed protocol of the study and the details regarding ‘Study population’ and ‘Data collection’ has been described previously.7 ,8 Briefly, 6301 of the study participants from Dongzhi community who received a baseline screening examination were invited for a follow-up visit in 2011, and 2901 (46%) of them responded. The non-responders did not differ substantially from the responders with respect to baseline characteristics (data not shown). This study was approved by the Institutional Review Boards from the Nanfang Hospital in Guangzhou and the Institute of Biomedicine in the Anhui Medical University. Written informed consent was obtained from each study participant.
Baseline data were collected by trained research staff according to protocols described previously.7 The question about physical activity was phrased as follows: “How do you describe your daily physical activity level?” and a choice of three responses: low, moderate and high was provided. The question about health status was phrased as follows: “How do you describe the change of your health status in the last year?” and a choice of three responses: obviously better, no obvious changes and obviously worse was provided.
Venous blood was drawn from the forearm of each participant in the fasting status. Serum and plasma were separated from blood cells in the field within 30 min and kept frozen at −20°C (no more than 8 h) until laboratory assays could be performed. Serum creatinine concentrations were determined using an enzymatic method (sarcosine oxidase-PAP) with a coefficient of variation of 1.1%. Fasting plasma glucose (FPG), total cholesterol (TC), triglyceride (TG) and high-density lipoprotein cholesterol (HDL-C) were measured on the Hitachi 7020 Automatic Analyzer.
Definition of RFD
eGFR was estimated using the following equation derived from the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI).9
RFD was defined according to the Kidney Disease: Improving Global Outcome (KDIGO) 201210 as follows: a drop in the GFR category (≥90 (G1), 60–89 (G2), 45–59 (G3a), 30–44 (G3b), 15–29 (G4), <15 (G5) mL/min/1.73 m2) accompanied by a 25% or greater drop in eGFR from baseline, or a sustained decline in eGFR of more than 5 mL/min/1.73 m2/year. The annual eGFR change was calculated as (eGFR at baseline−eGFR at revisit)/follow-up year.
Current smoking was defined as having smoked ≥10 packs in the past year. Current drinking was defined as drinking alcohol at least one time per week in the past year. Obesity was defined as a body mass index (BMI) of ≥23 kg/m2.11 Abdominal obesity was defined according to the guidelines of the International Diabetes Federation for Chinese populations as a waist circumference (WC) ≥90 cm for men or ≥80 cm for women.12 Blood pressure was categorised into three grades: normal, systolic blood pressure (SBP) <120 and diastolic blood pressure (DBP) <80 mm Hg; prehypertension, SBP 120–139 and/or DBP 80–89 mm Hg; hypertension, SBP ≥140 and/or DBP ≥90 mm Hg.13 Mean blood pressure (MBP) was calculated as one-third SBP plus two-thirds DBP. Pulse pressure (PP) was calculated as the difference between SBP and DBP.
Means (SD) or medians (25th centile, 75th centile) and proportions were calculated for population characteristics by sex. The differences in population characteristics were compared accordingly using Student t tests, signed rank tests or χ2 test. The adjusted whole and sex-specific ORs and 95% CI of developing RFD were determined from multivariable logistic regression models that included age group (40–50, 50–60 and ≥60 years), baseline eGFR (≥120, 90–120, <90 mL/min/1.73 m2), sex (men vs women), obesity or abdominal obesity, cigarette smoking, alcohol drinking, FPG (≥7.0 vs <7.0 mmol/L), TC (≥5.2 vs <5.2 mmol/L), TG (≥1.7 vs <1.7 mmol/L), HDL-C (<0.9 vs ≥0.9 mmol/L), self-reported health status change (better, no change, worse), education level (illiterate, primary level, elementary or higher levels), physical activity level (low, moderate, high) and blood pressure (normal, prehypertension, hypertension) or MBP (<80, 80–90 and ≥90 mm Hg) or PP (<40, 40–50 and ≥50 mm Hg). We also explored the possible interactive effects of sex with the proved risk factors in the above models on the risk of RFD, and further calculated the sex-specific adjusted ORs (95% CI) of developing RFD with the increasing numbers of proved risk factors in the above models by using the same methods. All of the analyses were performed using R software, V.2.14.1 (http://www.R-project.org).
Of the 2901 participants in the second examination cycle, those with known cancer (n=1), coronary heart disease or stroke (n=28), hypertension (n=126), diabetes (n=8), or with any missing data (n=45) about age, sex, height, weight, WC, smoking status, drinking status, SBP, DBP, FPG, TC, TG, HDL-C, self-reported health status change, education and physical activity levels, or with age <40 years (n=45) were excluded. Furthermore, we also excluded study participants with baseline eGFR <60 mL/min/1.73 m2 (n=130), resulting in a sample size of 2518 in the final analysis.
Population characteristics by sex are listed in table 1. Men had statistically significant higher age, WC, DBP, MBP, HDL-C, cigarette smoking, alcohol drinking, education levels and physical activity levels, but lower BMI, PP, TC, TG, eGFR and self-reported health status levels compared with women.
After an average 7.13-year follow-up (median: 7.08; 25th centile – 75th centile 7.03–7.13; 17 658 person-years), the incidence rate of RFD was 8.7% (women 7.4% and men 9.8%). Men had a faster eGFR decline than women (median decline: 1.90 vs 1.75 mL/min/1.73 m2/year, p=0.014, table 1).
In the multivariable logistic regression model, the ORs (95% CI) of developing RFD were 1.60 (1.01 to 2.54) for men versus women and 1.51 (1.09 to 2.08) for participants with obesity or abdominal obesity versus none (1.35 (0.85 to 2.14) for men and 1.65 (1.04 to 2.64) for women, p for interaction between obesity or abdominal obesity and sex=0.869). For other risk factors, sex-specific differences were observed. The ORs (95% CI) of developing RFD were 1.64 (1.02 to 2.63) and 2.05 (1.21 to 3.47), respectively, for participants with prehypertension and hypertension versus normal blood pressure (p values for interaction between prehypertension, hypertension and sex were 0.166 and 0.072, respectively), 2.63 (1.11 to 6.20) for participants with MBP ≥90 vs <80 mm Hg (p for interaction between MBP and sex=0.027), 2.00 (1.26 to 3.18) for participants with PP ≥50 vs <40 mm Hg (p for interaction between PP and sex=0.005), 2.65 (1.08 to 6.50) for participants with HDL-C <0.9 vs ≥0.9 mmol/L (p for interaction between HDL-C and sex=0.062), and 3.11 (1.59 to 6.10) for participants with low versus high physical activity levels (p for interaction between physical activity and sex=0.032) in men (table 2).
However, the ORs (95% CI) of developing RFD were 3.22 (1.22 to 8.46) for current smokers (p for interaction between current smoking and sex=0.005), and 2.57 (1.20 to 5.50) for participants with worse versus better self-reported health status (p for interaction between self-reported health status and sex=0.102) in women (table 2).
Furthermore, the ORs increased significantly with the increasing numbers of proved risk factors (prehypertension or hypertension, HDL-C <0.9 mmol/L, and low physical activity for men; and abdominal obesity, current smoking, and worse self-reported health status for women), particularly in men (table 3).
We obtained similar results when the participants with eGFR <90 mL/min/1.73 m2 (n=2302) or BMI <20 kg/m2 (n=1831) were excluded, or when the participants with eGFR <60 mL/min/1.73 m2 (n=2648) were included in our analysis (data not shown). We also repeated our analyses using the following outcomes: (1) a 25% or greater drop in eGFR from baseline; (2) a sustained decline in eGFR of more than 5 mL/min/1.73 m2/year; and came to the essentially similar results to our primary analyses.
Primary prevention, early detection and intervention in the relatively healthy population are the most cost-effective strategies to overcome the CKD epidemic. The prevalence of CKD varies significantly across areas and ethnicities.14 Therefore, designing appropriate health plans and preventive measures requires the gathering of necessary information on CKD from different geographical areas and ethnicities. However, to the best of our knowledge, only one previous prospective study among 1492 urban participants with an eGFR >30 mL/min/1.73 m2 had investigated the indicators of RFD in China, without considering some important lifestyle factors like physical activity levels, the possible sex difference and the relatively short follow-up (4 years) period.15 Our study was the first to longitudinally observe the risk factors for RFD in Chinese adults without known CVD, diabetes, hypertension, cancer and with normal renal function (eGFR ≥60 mL/min/1.73 m2).
Prehypertension, hypertension, higher MBP or PP were independent risk factors for RFD only in men in our present study. Consistently, Tohidi et al5 reported that hypertension and high normal blood pressure were significantly associated with incident CKD only among men in an Iranian cohort. The study by Kronborg et al6 also showed that a higher SBP in men was associated with a higher decrease of eGFR levels than in women (p for interaction between SBP and gender=0.019). Our results suggest that not only hypertension but also prehypertension as well as MBP and PP should be used to identify a more high-risk population for RFD in male Chinese adults.
Besides higher blood pressure, we observed that lower HDL-C and lower physical activity levels were major risk factors for RFD in men and current smoking was a major risk factor for RFD in women. In line with our findings, Tohidi et al5 found that dyslipidaemia was mainly associated with the risk of incident CKD in men (OR=1.38; 95% CI 0.93 to 2.03), but not in women (0.79; 0.60 to 1.05), during a mean follow-up of 9.9 years. Furthermore, despite the lower prevalence of smoking among women compared with men, the current female smokers showed an over fivefold risk for incident CKD.5 The effects of alcohol intake on the risk of RFD remain controversial. White et al16 reported that an alcohol intake of≥30 g/day increased the risk of albuminuria but reduced the risk of eGFR <60 mL/min/1.73 m2, whereas Yamagata et al17 reported that an alcohol intake did not have substantial effect on renal function. This study did not observe the significant relationship of alcohol drinking with RFD. However, in the PREVEND study, Halbesma et al4 reported that a higher WC and cholesterol/HDL ratio were associated with less RFD in males, whereas cholesterol/HDL ratio was associated with more RFD in females in the 5488 participants from the Netherlands during a mean follow-up of 6.5 years. Kronborg et al6 suggested that high alcohol consumption in men, current smoking and high physical activity in women predicted an increase in eGFR among 4441 general participants from Norway after a 7-year follow-up. Furthermore, Tohidi et al5 found that new diagnosed diabetes was an independent risk factor for CKD only among males; however, known diabetes was an independent risk factor for CKD among females in a community-based cohort of the Middle East population, during a mean follow-up of 9.9 years. The reasons for the discrepancies were not fully understood, but may possibly be explained by the various ethnicities, culture, lifestyles and follow-up time. Future studies are needed to further investigate these issues.
Self-reported health represents the physical, emotional and social aspects of health and well-being.18 Patients with poor self-reported health had a significantly increased risk of mortality even after controlling for demographic and clinical confounders in incident dialysis patients.19 This study further showed that worse self-reported health may also be a predictor for RFD in women. The lack of significant association between self-reported health and RFD among men might be related to lack of power. These results imply that we should possibly transcend the only traditional ‘sick’ condition control and focus more on the emotional or psychosocial factors in CKD prevention. However, there were similar annual eGFR change levels in women stratified by self-reported health status (median: 1.84, 1.81 and 1.89 mL/min/1.73 m2/year, respectively, for better, no change and worse self-reported health status). The possible explanation was that self-reported health was associated with both the rapid RFD and the increase in eGFR. As previously reported,3 there was an increase in mortality risk with an increase in estimated GFR, particularly among individuals with higher first estimated GFR (≥60 mL/min/1.73 m2). We would evaluate the factors associated with the obvious increase in eGFR in our future studies.
Just as in a previous study,20 age was not an independent risk factor for RFD in this study. The increasing incidence of RFD in older individuals might possibly result from an increase in age-related risk factors for the development of RFD. We also did not have the detailed diet information at baseline. However, the information of obesity or abdominal obesity and blood lipid may partly represent the status of red meat or fruit and vegetable consumption.21 BMI and WC represent different aspects of body composition: BMI is a surrogate of overall adiposity while WC is a surrogate of central adiposity. Our results with obesity or abdominal obesity in the model (adjusted OR 1.65; 95% CI 1.04 to 2.64) suggest an independent effect of obesity or abdominal obesity on RFD in women. The other analysis: (1) obesity only (1.61; 1.00 to 2.59); and (2) abdominal obesity only (1.76; 1.06 to 2.93) in the same models yielded the similar results. In the stratified analysis, the increased (not significant) risk of RFD associated with abdominal obesity was observed in participants with normal weight (1.55; 0.50 to 4.76) or obesity (1.45; 0.66 to 3.19). However, the increased risk associated with obesity was only observed in participants with normal WC (1.43; 0.72 to 2.83; abdominal obesity: 0.89; 0.26 to 0.98). Our results suggested that control of obesity and abdominal obesity was required to reduce the risk of RFD; however, abdominal obesity may possibly be a more important risk factor for RFD. Consistently, Kim et al22 also reported that percentage body fat was associated with eGFR changes, even in participants with a normal weight.
The prevalence of at least one significant risk factor for men (prehypertension or hypertension, HDL-C <0.9 mmol/L and low physical activity) and women (obesity or abdominal obesity, current smoking and worse self-reported health status) was 69.9% and 56.9%, respectively. In addition, there was a detrimentally additive effect among these risk factors. Our findings underscore the urgent need to take sex-specific measures to reduce the huge CKD burden in China.
Both previous animal and human studies found that male gender was associated with a more rapid rate of progression and a worse renal outcome in participants with CKD.23 ,24 This study also showed that men had a significantly faster eGFR decline and a higher RFD risk, independent of other traditional risk factors. The underlying mechanisms for this sex difference potentially involved differences in glomerular structure, glomerular haemodynamics, variations in the production and activity of local cytokines and hormones, and the direct effect of sex hormones on kidney cells.24 The reasons that males have greater kidney bulk and weight than do females are unclear, but may be related to the body surface area and the effects of androgens.25 ,26 However, we did not have the detailed information for hormones and only adults aged 40 years or more were included in this study. A further understanding of the sex differences would provide us with some new strategies for controlling CKD.
We excluded participants with known CVD, diabetes, hypertension, cancer and CKD (eGFR <60 mL/min/1.73 m2) from our analyses to exclude the possibility of confounding due to concomitant diseases, medications or lifestyle changes. Our study also has several limitations. First, although a lot of what might seem to be healthy participants may have underlying pathology (which could not be adjusted in our present analysis) due to lack of physician access in rural communities, we conducted the study in a relatively healthy Chinese population. In fact, this was the only study to examine the sex-specific risk factors for RFD in participants with normal renal function (eGFR ≥60 mL/min/1.73 m2). The mean eGFR level was 108.3 mL/min/1.73 m2, which was consistent with a previous report conducted among healthy Asians without kidney disease.27 Hence, caution is needed in generalising our findings to other populations or ethnicities, particularly participants with lower eGFR levels or obvious kidney disease. Second, just like most studies of CKD, epidemiological or interventional, single serum creatinine measurement was used in our study. Third, albuminuria was also an important marker of kidney damage.28 We did not have urinary albumin data at baseline and therefore were not able to perform the analysis on albuminuria. However, our current analysis mainly aimed to examine the risk factors for RFD. eGFR was a widely used indicator of renal function in clinical and public health practice. Declines in eGFR were strongly and consistently associated with the risk of ESRD and mortality independent of albuminuria.3 Nevertheless, future studies with both albuminuria and eGFR measurements should be performed to evaluate the possible modifying effect of albuminuria on our results. Furthermore, the questions included in the standardised questionnaire used to collect data on lifestyle habits in our study are subjective. Future studies designed to examine the association between the objectively measured lifestyle habits suggested in our studies and RFD are needed to further support our results.
In conclusion, we observed the sex-specific risk factors for RFD in Chinese adults with normal kidney function and there was a detrimentally additive effect among these significant risk factors. Screening to detect RFD and early intervention to modify these risk factors should be considered in this relatively healthy population. However, more studies are needed to further support our results, and to determine whether implementation of these measures could result in a reduction of RFD and CKD incidence in this population.
What is already known on this subject?
There is increasing but inconclusive evidence of sex-specific differences in the association of some risk factors with the risk of renal function decline, particularly in participants with normal renal function.
What this study adds?
Our study was the first to longitudinally observe the risk factors for renal function decline (RFD) in Chinese adults with normal renal function.
Our findings suggested that sex-specific risk factors should be considered in prevention of RFD in Chinese adults with normal kidney function.
Contributors XipX, XQ, YL, DX, GT, BW, XW, XiX and FH were involved in design of the study. XipX, XQ, YW, YL, DX, GT and BW were involved in conduct of the study. XipX, XQ, YW, YL and XiX were involved in data collection and analysis. XipX, XQ, YW, YL, DX, GT, BW, XW, XiX and FH were involved in data interpretation and manuscript writing. All authors read and approved the final manuscript.
Funding The study was supported by the Major State Basic Research Development Program of China (973 program) (No. 2012CB517703), the Public Welfare and Health Sector Research Project (201002010) and Major Scientific and Technological Planning Project of Guangzhou City (2010U1-E00821), and the National Nature and Science Grant (No. 81202280, No. 81402735, Grant No. 81473052, Grant No. 81441091).
Competing interests None.
Ethics approval This study was approved by the Institutional Review Boards from the Nanfang Hospital in Guangzhou and the Institute of Biomedicine in the Anhui Medical University.
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.