Background
Recently, with the gradual change of modern lifestyles and diets, the incidence of hyperuricemia is increasing. Uric acid (UA) is preponderantly excreted by the kidney; hence, aberrant production or clearance by the kidney, such as overproduction or underexcretion, can increase the UA levels [
1]. A dramatic rise in the serum UA levels (SUA) was originally proposed as the cause of gout [
2]. Based on observations of gout patients, Mahomed et al. hypothesized that hyperuricemia is a possible mediator of hypertension [
3]. Subsequently, Haig suggested that UA could lead to many diseases in addition to gout, such as rheumatism, hypertension, diabetes, and chronic kidney disease (CKD) [
4]. Since then, numerous epidemiologic studies have successively indicated that there is a relationship between elevated UA levels and metabolic syndrome [
5], renal disease [
5,
6], hypertension [
7], and cardiovascular disease (CVD) [
6,
8].
CKD has emerged as a primary public health problem worldwide [
9], and the epidemic prevalence has doubled in the past several decades [
10], particularly in China [
11] and other developing countries [
10]. CKD can lead to not only end-stage renal disease (ESRD), but also complications related to renal impairment and an increased risk of CVD [
9].
As the diagnostic techniques and aggressive treatment strategies have developed, renewed attention has been given to elevated SUA levels. The relationship between the UA level and development of CKD has been supported by expanding epidemiologic and experimental evidence. Researchers have investigated the role of UA not only as a potential marker of renal dysfunction [
6,
12‐
14], but also as a significant pathogenic factor that is involved in the development of renal disease [
12,
15‐
18]. Basic experimental studies have confirmed that the potential mechanisms of renal injury due to hyperuricemia induce inflammation, afferent arteriopathy [
16], and endothelial dysfunction [
18]; activate the renin-angiotensin-aldosterone system (RAAS) [
17] and cyclooxygenase-2 (COX-2) expression; and impair oxidative metabolism [
12], among others. In contrast, many large epidemiologic studies have demonstrated that elevated SUA levels were an independent risk factor for the worsening of renal function and incident CKD in the general population [
19,
20], after adjusting for age, sex, race, diabetes, systolic blood pressure (SBP), hypertension, lipids, history of underlying disease, baseline estimated glomerular filtration rate (eGFR), living habits, and other factors [
19‐
22]. Furthermore, a prospective cohort study of healthy people suggested that there was a U-shaped association between UA and the loss of kidney function. This means that both low and high UA levels may predict a decline in kidney function [
23]. However, many cohort studies have shown conflicting results, indicating that there was no significant association between an increased SUA level and CKD progression [
24‐
26], especially after adjustment for confounders.
Therefore, to avoid the instability of a single test, which might not be sufficient for identifying patients at risk, we aimed to investigate whether the time-mean SUA value indicates the risk of CKD, and explored the association of the baseline and time-mean SUA levels with kidney function decline and incident CKD in an ostensibly healthy population.
Discussion
We demonstrated that high levels of SUA are associated with the occurrence of CKD, while low SUA levels are not. Importantly, the major results of our study demonstrate that high SUA levels, particularly the time-mean SUA values, indicate the risk of renal progression and renal dysfunction after adjustment for confounding variables, and this association was even observed in the normal range of SUA levels.
It is controversial whether SUA plays a causal role in the progression of CKD or if it is merely a marker of renal damage. Many observational studies indicated that hyperuricemia is an independent risk factor for the development and progression of renal disease in healthy individuals [
19‐
22]. However, other prospective observational studies have produced contrary results; they did not show that the SUA level had a positive effect on the incidence of CKD in Japanese patients [
24‐
26]. On the other hand, Sturm et al. revealed that hyperuricemia predicts the progression of CKD, but only before correction. After adjusting for baseline kidney function parameters such as baseline eGFR and proteinuria, they found that hyperuricemia no longer acted as a risk factor for the progression of CKD [
25].
Our findings suggest that the risk of renal function reduction might increase with increased SUA levels, especially the time-mean SUA level, even those within the normal range. First, in contrast with other studies on patients with CKD in other countries, in Chinese patients, the influences of high SUA levels on the natural history of renal function have been less examined. Our results show that in subjects who undergo annual health check-ups, the increase in the SUA level was associated with a slow decline in the eGFR, but there was a high incidence of renal insufficiency. The relationship between the SUA levels and eGFR that was observed in our study was similar to the relationship reported by a multi-center study in Japan, which included 141,514 subjects without renal insufficiency at baseline [
29]. It was not similar to the relationship observed in a study by Kanda et al. in Japan [
23] suggesting that the UA level has a U-shaped relationship with the loss of renal function in men, indicating that both low and high UA levels were associated with a decline in eGFR. We believe that the difference between our results and theirs [
23] might be because their data were examined every 3 years, and the mean baseline UA levels in their cohort study were lower than in ours.
Second, our results revealed that high time-mean SUA levels affected the likelihood of new-onset CKD more than low levels after adjustment for confounders, while the association with baseline SUA levels was relatively milder. This is consistent with previous studies showing that elevated serum UA levels were independently associated with an increased risk of the development of CKD [
20,
22,
26]. We further examined the association between the time-mean SUA values and CKD after separately evaluating participants aged > 50 years and ≤ 50 years. In both sex models, the association was present in male and female subjects > 50 years old, but absent in women ≤50 years.
Third, our analyses showed that the statistical difference was subdued in the association between the baseline SUA and incidence of CKD, but they were still statistically significant in the time-mean UA value. Participants in the highest quartile of the time-mean SUA value (> 6.5 mg/dl) had a risk of developing CKD that was more than three times higher than in those in the lowest quartile. Consistently, two studies clarified that increased SUA accelerates progression to ESRD, and indicated that the target SUA level should be less than 6.5 mg/dl [
30,
31]. Similarly, Rudolf et al. found that the risk of new-onset CKD increased roughly linearly with the UA level to a level of approximately 6 to 7 mg/dl in women and 7 to 8 mg/dl in men. Above these levels, the associated risk increased rapidly [
22]. Our results are consistent with those of a recent meta-analysis that included 15 cohort studies and demonstrated that there was a positive association between the SUA levels and risk of CKD in middle-aged patients, independent of established metabolic risk factors [
32].
Many studies have clarified the mechanisms by which hyperuricemia leads to loss of kidney function. Animal experiments found that hyperuricemia may induce endothelial cell dysfunction and inhibit the generation of nitric oxide [
12,
18]. In addition, studies have demonstrated that hyperuricemia induced arteriolopathy of the preglomerular vessels, which impairs the autoregulatory response of the afferent arterioles. Simultaneously, vascular wall thickening such as platelet adhesiveness and disturbed hemorheology [
33] can result in ischemia that is induced by lumen obliteration [
17]. Renal hypoperfusion is a potent stimulus of vasoactive and inflammatory mediators, ultimately leading to tubulointerstitial inflammation and fibrosis [
17]. Other animal experiments demonstrated that activation of the RAAS and COX-2 systems could be mediated by upregulating the angiotensin-1 receptors on the vascular smooth muscle cells [
12,
17].
The major strengths of this study were that the participants’ serum creatinine and serum UA levels were not only measured at baseline, but were also obtained every year during the subsequent follow-up period. Therefore, we could evaluate the accurate date of the onset of CKD and explore the independent effect of the time-mean SUA levels on the renal outcome. Moreover, we could avoid instability due to short-term fluctuation. Second, we analyzed the serum UA (mean ± SD: 5.6 mg/dl ± 1.3) in this cohort, which is considered representative of the Chinese general population. Third, we used the recently developed Modification of Diet in Renal Disease equation, which is known to be more accurate for Chinese people than other methods, and we adopted an eGFR of < 60 ml/min/1.73 m2 as the criterion to determine CKD.
Nevertheless, our study has some limitations that should be considered. First, the major limitation of our study is that the subject selection process was not entirely random. The participants were relatively healthy individuals who actively paid attention to their health status; therefore, a self-selected bias cannot be excluded in this cohort. Second, we were unable to collect data on medication history, such as UA-lowering medicines, diuretics, and antihypertensive drugs. In addition, we did not have information on lifestyle-related factors such as dietary habits, smoking, alcohol consumption, and exercise. Therefore, this study did not eliminate all the factors that might impact changes in the SUA levels and inhibit the progression of CKD.