Uric acid level and kidney function: a cross-sectional study of the Korean national health and nutrition examination survey (2016–2017)

Kidney disease is expected to become the fifth leading cause of premature death globally by 2040. Uric acid level is a risk factor for kidney disease. The current study aims to investigate the association between uric acid levels and kidney function in the Korean population. The data of 11,042 participants of the 2016–2017 Korea National Health and Nutrition Examination Survey were analysed. The estimated glomerular filtration rate was calculated using the modification of diet in renal disease formula for Koreans. For each sex, uric acid levels were divided into five subsequent categories of increasing levels (Q1, Q2, Q3, Q4, and hyperuricemia). The association between uric acid level and kidney function was investigated using multiple logistic regression. The results showed that the higher the uric acid levels, the greater the odds of reduced kidney function in both sexes. In men, the adjusted odds ratios (95% confidence intervals) for reduced eGFR comparing the hyperuricemia group to the lowest serum uric acid quartile was 5.55 (3.27–9.44), and in women, the odds ratios (95% confidence intervals) was 7.52 (4.39–12.87). Normal weight or underweight in men and overweight in women, as well as diabetes mellitus, hypertension, and physical inactivity were highly associated with reduced kidney function. Our study revealed a dose–response relationship between uric acid levels and kidney function. Therefore, high uric acid level should be considered as a factor that is potentially related to kidney dysfunction in the Korean population.


Materials and methods
Data collection and study population. Data were derived from the Korea National Health and Nutrition Examination Survey (2016-2017) 18 . KNHANES is a cross-sectional study design of nationally representative to evaluate the health and nutritional status of the Korean population aged 1 year and over. This survey is conducted annually by the Korean Centers for Disease Control and Prevention (KCDC). KNHANES employs a stratified and multistage cluster sampling design based on geographic area, gender, and age, to select household units. KNHANES contains reliable data on the general population of Korea, and is a valuable resource for development and evaluation health policies and programs in Korea. The initial study population comprised 16,277 individuals. Among them, 16,232 subjects were selected, and those who were diagnosed with kidney failure by doctor were excluded. In addition, the subjects with missing data on uric acid levels, sex, age, socioeconomic status, or health-related factors were excluded from the study, resulting in a final sample of 11,042 individuals for analysis. KNHANES data is publicly accessible and ethical approval is not required for the use of the data. In addition, as the respondent's information is completely anonymous, there is no need for prior consent for research purposes.

Measurement of kidney function using the eGFR for the Korean population.
The main objective of this study was assessment of kidney function. GFR is an essential aspect of kidney function evaluation. This measurement determines the level of creatinine in the blood and calculates a kidney function score that indicates how well the kidneys are functioning. Glomerular filtration rate indicates the amount of blood filtered per minute by the glomeruli. If kidney function decreases due to injury or disease, the filtration rate decreases, and waste products begin to accumulate in the blood. Clinically, the eGFR is mainly used to detect kidney status. eGFR is calculated on the basis of a serum creatinine test. Creatinine is a muscle waste product that is filtered out of the blood by the kidneys and is released into the urine at a relatively constant rate.
The eGFR was calculated using the modification of diet in renal disease (MDRD) formula. In 2010, a Korean MDRD formula was developed, which was reported to measure the glomerular filtration rate of Koreans more accurately than the conventional MDRD formula. The Korean MDRD formula for the eGFR is as follows: 107.904 × (Creatinine in mg/dL) −1.009 × (age) −0.02 × (0.667 if female) 19 . Therefore, in this study, kidney function was confirmed using the Korean MDRD formula. The normal range of glomerular filtration rate is about 90-120 mL per minute. A reduction of the GFR can be interpreted as impaired kidney function. According to the eGFR formula, participants with a GFR of less than 90 mL were considered to have reduced kidney function. In order to identify malfunction of kidney function, the dependent variable was designed in the form of a binary division with a cut-off of 90 mL (i.e., decreased function group: < 90 mL/min/1.73 m 2 , normal group: ≥ 90 mL/ min/1.73 m 2 ) 2 .
Uric acid levels. The main exposure of interest was serum uric acid levels (SUA). Venous blood samples were collected in the morning after an overnight fast. In the KNHANES, SUA was measured by means of a Hitachi automatic analyzer 7600-210 using a colorimetric enzymatic method. SUA levels were then divided into five categories in both sexes: one group for hyperuricemia (> 7.00 mg/dL for men; > 6.00 mg/dL for women 20,21 ), and four groups (Q1, Q2, Q3, and Q4) within the reference range according to sex-specific quartiles. The SUA quartiles were as follows: < 4.80 mg/dL, 4.80-5.49 mg/dL, 5.50-6.09 mg/dL, 6.10-6.99 mg/dL for men; and < 3.70 mg/dL, 3.70-4.19 mg/dL, 4.20-4.79 mg/dL, 4.80-5.99 mg/dL for women. The groups under 4.80 mg/dL (men) and 3.70 mg/dL (women) were set as the reference groups.
Covariates. The following covariates were included in the fully adjusted models because they could be associated with kidney function and uric acid level: age (< 40, 40-49, 50-59, 60-69, ≥ 70), household income group (low, medium-low, medium-high, or high), educational level (high school or under, university or above), region (metropolitan or rural), and occupational categories (white collar, pink collar, blue collar, or unemployed). Household income groups classified as quartiles were calculated by dividing household income by the square root of the number of household members, a standard method recommended by the Organisation for Economic Cooperation and Development 22 23 . We restructured the classification into four categories: white (office work), pink (sales and service), blue (agriculture, forestry, fishery, and armed forces occupations), and unemployed. Health-related covariates included body mass index (normal or underweight, overweight, or obese), waist circumference (normal circumference or abdominal obesity), diabetes mellitus, hypertension, dyslipidaemia, smoking status (non-smoker, ex-smoker, or current smoker), frequency of drinking (never, occasionally, or frequently), physical activity (active or inactive). Waist circumference was divided using the Korean abdominal obesity criteria, which is 90 cm for men and 85 cm for women 24 . The diagnoses of diabetes mellitus, hypertension, and dyslipidaemia were confirmed by an experienced physician. Physical activity was assessed by the ability of the subject to engage in a moderate-intensity activity for more than 150 min, a high intensity activity for more than 75 min, or a combination of both moderate and high intensity-activity (1 min of high intensity and 2 min of moderate intensity) per week 25 .

Statistical analysis.
Owing to the considerable sex differences in physical functions such as the capacity of female hormones to effectively reduce uric acid levels, all analyses have been stratified by sex. Sampling weights were applied in all data analyses to perform multistage stratified probability sampling of KNHANES. Descriptive analysis was performed to examine the distribution of the general characteristics of the study population. We calculated the frequency and percentages for each variable through the chi-square test. The statistical significance level was set as p-value of lower than 0.05. To identify the association between uric acid levels and kidney function, a multiple logistic regression analysis was performed after adjusting sociodemographic and health-related covariates. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated in order to perform comparisons between subjects with uric acid levels below 4.80 mg/dL (males) or 3.70 mg/dL (females) and subjects with different levels of uric acid (males: 4.80-5.49 mg/dL, 5.50-6.09 mg/dL, 6.10-6.99 mg/dL, over 7.00 mg/dL; females: 3.70-4.19 mg/dL, 4.20-4.79 mg/dL, 4.80-5.99 mg/dL, over 6.00 mg/dL). Subgroup analysis was performed by body mass index (BMI), waist circumference, age, previously diagnosed diabetes, previously diagnosed hypertension, previously diagnosed dyslipidaemia. In addition, multicollinearity was tested using the variance inflation factors. All statistical analyses were performed using SAS 9.4 software (SAS Institute, Cary, NC, USA). Table 1 presents the general characteristics of the gender-stratified study population. Overall, 922 (19.0%) of the 4,848 men and 403 (6.5%) of the 6194 women included in the study were considered hyperuricemic respectively. In general, among both men and women, higher uric acid levels were associated with lower kidney function (p < 0.001 for both men and women). Specifically, the percentages of decreased kidney function among men were 6.6%, 9.3%, 10.8%, 15.4%, and 24.8% in the < 4.80, 4.80-5.49, 5.50-6.09, 6.10-6.99, and ≥ 7.00 mg/dL groups, respectively, while among women, the proportions of decreased kidney function were 19.8, 25.9, 35.5, 43.1, and 61.8% in the < 3.70, 3.70-4.19, 4.20-4.79, 4.80-5.99, and ≥ 6.00 mg/dL groups, respectively. Table 2 shows the logistic regression analysis results for both men and women, adjusted for all covariates. We observed a dose-response relationship between uric acid levels and kidney function in both men and women. In particular, compared with the reference group, the odds ratios (95% CIs) for kidney dysfunction in men were  Fig. 1A and B present the results of the subgroup analysis stratified by gender. For BMI, both genders maintained the dose-response relationship in all BMI categories. Among men, the ORs of the normal or underweight group were significantly higher than those of the other groups. In particular, the OR of the hyperuricemia group was 9.86 times higher than that of normal subjects. In contrast, among women, the OR of the hyperuricemia subjects in the overweight group was 6.94 times higher than that of normal subjects.

Results
In the subgroup analysis according to waist circumference and BMI, a dose-response relationship was observed in both the normal and the abdominal obesity groups. Among women, those who had abdominal obesity and hyperuricemia were 6.48 times more likely than normal controls to show decreased kidney function. The likelihood of decreased kidney function in both men and women increased with advancing age. Men who were previously diagnosed with diabetes or hypertension showed significantly higher ORs than did normal subjects; in particular, men with hyperuricemia were 7.71, and 9.79 times higher than the reference group. Women who were previously diagnosed with diabetes, hypertension, and dyslipidaemia had a dose-response relationship between uric acid levels and decreased kidney function. Particularly, in the hyperuricemia group, ORs were the highest at 9.13, 9.89, and 10.88 times. Moreover, subjects with limited physical activity were more likely to have decreased kidney function.

Discussion
The purpose of this study was to examine the association between uric acid levels and kidney function in the Korean population using the Korean MDRD formula using representative KNHANES data. We also conducted a subgroup analysis according to BMI, waist circumference, diabetes, hypertension, dyslipidemia, and physical activity, which are factors related to kidney disease.
In this study, we measured the GFR using the MDRD formula for the Korean population to evaluate kidney function. While the direct measurement of GFR is considered to be the most accurate way to detect changes in kidney function, direct measurement requires special skills and involves complex measurements. Therefore, the eGFR is often used in clinical practice. The MDRD formula was derived from studies on the relationship between protein intake and kidney failure progression in patients with CKD in the United States. However, since the MDRD formula was derived mainly from research conducted in populations of European descent, the accuracy of the formula was not verified for other races 26,27 . In 2010, a Korean MDRD formula was developed, allowing for more accurate GFR measurement in Koreans than that with the conventional MDRD formula 19 .
We observed a dose-response relationship between uric acid levels and kidney function in both sexes. Our results indicate that an increasing uric acid level is significantly associated with an increased odds ratio of decreased kidney function. These results are similar to those of earlier studies confirming that high uric acid levels are a risk factor for kidney dysfunction 15,[27][28][29] . Animal studies have also shown that elevated blood uric acid levels can cause kidney damage 26,30 .  Although there is controversy about the association between uric acid levels and CKD, there are several possible mechanisms by which higher uric acid levels could play a role in the development of CKD. According to previous studies, increased levels of uric acid, especially hyperuricemia, can contribute to endothelial dysfunction by inducing antiproliferative effects on the endothelium and inhibiting the production of nitric oxide [31][32][33] . Over time, damage to the lining of the blood vessels can lead to CKD 34 . A cohort study indicated that early treatment of hyperuricemia could potentially control CKD 35 .
Women had a slightly higher risk of kidney dysfunction than men, which is consistent with the findings of previous studies 20, 36,37 . However, other studies have reported that gout and hyperuricemia affect men more commonly 38 . This is possibly related to the female hormone estrogen, which increases uric acid clearance; indeed, premenopausal women have lower uric acid levels than men, and uric acid levels in women start to rise in the fifth decade of life [38][39][40] . In particular, kidney function was significantly reduced in both men and women in the hyperuricemia group. Patients with CKD often have hyperuricemia 29,41,42 . A recent study revealed that even asymptomatic hyperuricemia that does not develop into gout can affect kidney function unless it is treated with uric acid-lowering drugs 43 .
BMI and waist circumference are known to affect uric acid levels 44 . Our subgroup analysis revealed that women with overweight and abdominal obesity have a more pronounced association with kidney disease. Previous studies have also shown that people with obesity have a higher risk of kidney disease 45 . In addition, uric acid levels and the likelihood of kidney disease development increased with age, as in previous research studies 46 . Diabetes, hypertension, and dyslipidemia are also powerful risk factors for kidney disease 6,34,47 . In addition, physical activity was found to be positively associated with kidney function 48,49 .
This study has several limitations that should be considered when interpreting the results. First, cross-sectional data cannot be used to infer any causal relationships. However, unlike a previous cross-sectional study that utilized hospitalized patients 50 , our study explored the relationship between uric acid level and kidney function using nationally representative samples of South Korea. Further research is still needed to characterize the longitudinal relationship between uric acid levels and kidney diseases in the South Korean population. Second, factors related to health and behavior, including previously diagnosed diseases, were self-reported. Thus, the possibility of recall bias cannot be discarded, despite the efforts of the surveying agency to reduce such bias. Third, although we adjusted for covariates related to uric acid levels and kidney function, the confounding effect cannot be completely excluded because other confounding variables might exist. Particularly, we did not consider the use of uric acid-lowering drugs such as allopurinol and febuxostat. Fourth, additional formulas, such as the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation, are available for measuring the eGFR. However, this study did not compare the differences in results using different formulas.
Despite these limitations, our research has several strengths. The use of nationally representative data allows our results to be generalized to the general adult population of South Korea. To the best of our knowledge, few studies has been conducted to investigate the association between uric acid levels and kidney function in the Korean population. Thus, this study is meaningful in that it investigated the correlation between uric acid levels and kidney function according to sex. Furthermore, blood samples were collected using standardized laboratory procedures, thus ensuring an accurate estimate of serum creatinine and uric acid levels.

Conclusion
Our study demonstrated a significant inverse association between high uric acid level and kidney function in the South Korean population. In particular, in the analysis stratified by obesity status, the association was stronger in BMI ≥ 23 group and abdominally obese participants compared with the normal group. Our research shows that uric acid level has a potential role in developing kidney disease. This could help health policymakers and professionals to implement strategies of prevention and intervention for CKD, such as blood screening for patients at a high risk of CKD.