Association between Serum Bilirubin and Estimated Glomerular Filtration Rate among Diabetic Patients

The subjects comprised 230 men aged 77 ± 10 (range: 50–100) years and 279 women aged 81 ± 10 (50–101) years that visited the medical department. We examined the relationship between increased serum bilirubin and renal function evaluated by estimated glomerular filtration rate (eGFR) using CKD-EPI equations modified by a Japanese coefficient. Compared with the fourth quartile in serum bilirubin (1.01–1.97 mg/dL), the nonadjusted, age and gender-adjusted, and multivariate-adjusted odds ratios {95% confidence interval (CI)} of eGFR <60 mL/min/1.73 m2 for the first quartile in serum bilirubin (0.13–0.50 mg/dL) were 2.08 (1.25–3.44), 1.82 (1.07–3.09), and 1.53 (0.83–2.81), respectively. Moreover, compared with the fourth quartile, nonadjusted, age and gender-adjusted, and multivariate-adjusted odds ratios (95% CI) of eGFR <45 mL/min/1.73 m2 for the first quartile were 3.50 (1.95–6.23), 3.12 (1.72–5.65), and 3.53 (1.71–7.26), respectively. The data were further stratified by gender, age, medication (antihypertensive, antidyslipidemic, and antidiabetic agents), and prevalence of cardiovascular disease (CVD). The standardized coefficients for eGFR were significant in all the subgroups other than the prevalence of CVD, and there were significant interactions between the two groups regarding CVD. Our data demonstrated an independent positive association between serum bilirubin and eGFR among diabetic patients.


Introduction
Serum bilirubin may protect against inflammation, cardiovascular disease (CVD), and all-cause mortality in adults [1,2]. Moreover, current evidences demonstrate that mildly elevated serum bilirubin may confer potent antioxidant properties, as indicated by its ability to scavenge peroxyl radicals and to inhibit oxidation of low-density lipoprotein (LDL) derived lipids [3,4]. Lots of studies have shown a positive relationship between serum bilirubin and estimated glomerular filtration rate (eGFR) [5][6][7][8][9], showing that serum bilirubin has a potential renoprotective effect. We also demonstrated an independent positive association between serum bilirubin and eGFR in both genders among elderly persons [10]. Therefore, it is reasonable to speculate that serum bilirubin levels may be negatively correlated with diabetic nephropathy and renal function among diabetic patients.
Several cross-sectional studies have shown that low serum bilirubin levels were significantly associated with decreased eGFR, and negatively associated with diabetic nephropathy in a hospital-based sample of diabetic patients [6,7]. In a cohort of Japanese type 2 diabetic patients, Mashitani et al. [9,11] demonstrated that serum bilirubin levels were prospectively associated with diabetic nephropathy progression, independent of possible confounders. In contrast, Targher et al. [12] found that serum bilirubin was negatively associated with eGFR, considering serum bilirubin as a renal risk factor. Thus, a relationship between serum bilirubin and renal function remains controversial.
We evaluated the relationship of serum bilirubin with confounding risk factors such as renal function, as well as hypertension, hyperglycemia, and lipids, using crosssectional data from the Nomura study [10].

Subjects.
Patients for this investigation were recruited among consecutive diabetic patients aged ≥50 years that 2 International Scholarly Research Notices visited the Medical Department of Seiyo Municipal Nomura Hospital. Patients with serum bilirubin > 2.0 mg/dL and severe cardiorenal (e.g., Gilbert's syndrome) or nutritional disorders that would affect blood pressure, lipid, and glucose metabolism were excluded. Thus, 509 persons were enrolled in the study. All procedures were approved by the Ethics Committee of Seiyo Municipal Nomura Hospital, and written informed consent was obtained from each patient.

Relationship of Serum Bilirubin and eGFR within Selected
Subgroups. Next, to control potential confounding factors, the data were further stratified by gender, age, medication (antihypertensive, antidyslipidemic, and antidiabetic agents), serum uric acid (first-second tertiles, third tertile), and  : standardized coefficient. Medication included antihypertensive, antidyslipidemic, and antidiabetic agents. § Adjusted for all confounding factors in the stepwise method in Table 2 by multiple linear regression analysis. prevalence of CVD (Table 4). The standardized coefficients for eGFR were significant in all subgroups other than the prevalence of CVD, and there were significant interactions only between the two groups regarding CVD.

Discussion
To examine the possible contribution of decreased serum bilirubin to renal dysfunction among diabetic persons, we studied the relationship between potential confounding risk factors including serum bilirubin and eGFR. This study showed a graded decrease in eGFR with decreasing serum bilirubin. Individuals with hypobilirubinemia (first quartile of serum bilirubin, <0.50 mg/dL) showed increased risk for stage 3b CKD (eGFR < 45 mL/min/1.73 m 2 ). Moreover, the strength of serum bilirubin level as an independent determinant of eGFR was similar to those of known factors such as gender, age, prevalence of antihypertensive medication, HDL-C, and serum uric acid. To our knowledge, few epidemiologic studies have quantified the link between decreased serum bilirubin and renal dysfunction in diabetic patients. Thus, we think that serum bilirubin levels may be utilized as a provisional new confounding factor of diabetic nephropathy that can be measured easily and applied in medical practice.
Several studies have shown that decreased serum bilirubin is a risk factor for the development of CKD among type 2 diabetic individuals. Inoguchi et al. [15] showed a lower prevalence of vascular complications as well as reduced markers of oxidative stress and inflammation in patients with Gilbert's syndrome, which is a congenital hyperbilirubinemia, and diabetes. A community-based cross-sectional study in Korea [7] found that total serum bilirubin levels were negatively correlated with 24-hour proteinuria and positively associated with eGFR after adjusting for potential confounding factors in 612 diabetic patients. Fukui et al. [6] found that serum bilirubin level was independently and negatively associated with albuminuria in a hospitalbased cross-sectional study in 633 Japanese type 2 diabetic patients. In addition, it was shown that serum bilirubin levels were higher in patients without diabetic nephropathy than in those with diabetic nephropathy. In a longitudinal cohort study of 12,823 Korean male workers without CKD or proteinuria at baseline, higher serum direct bilirubin levels were significantly associated with a lower risk of developing CKD (eGFR, <60 mL/min/1.73 m 2 ), even after adjusting for potential confounding factors [5]. In the single-center longitudinal observational cohort study of type 2 diabetic patients, Toya et al. [11] found that higher serum bilirubin levels, within the normal range, were associated with a lower risk of progression from microalbuminuria to macroalbuminuria. In a hospital-based study of 2,678 US diabetic outpatients (mean age: 55±18 years), Targher et al. [12] found International Scholarly Research Notices 5 that serum bilirubin levels were negatively associated with eGFR. However, in that study, no adjustment was made for potential confounding factors. Longitudinal data from 2,511 type 2 diabetic Japanese patients showed that multivariableadjusted odds ratios for progression from microalbuminuria to macroalbuminuria for the second, third, and fourth quartiles of serum bilirubin levels were 0.89 (95% CI 0.49-1.58), 0.93 (0.47-1.83), and 0.33 (0.13-0.84), respectively. However, this trend disappeared after adjustment for hemoglobin level [9]. In our hospital-based sample of 509 individuals, we found that decreased serum bilirubin levels were significantly associated with decreased eGFR, and multivariate adjustedodds ratio of hypobilirubinemia (0.13-0.50 mg/dL) for stage 3b (eGFR < 45 mL/min/1.73 m 2 ) was 3.53 (1.71-7.26).
The mechanism by which serum bilirubin level is associated with a lower risk of CKD is not completely understood. Bilirubin has been described as the most powerful endogenous antioxidant substance in vitro [16] when acting alone and complexed with serum albumin to serve as a superoxide scavenger and peroxyl radical trapping antioxidant [17]. Hyperglycemia causes mitochondrial superoxide overproduction in vascular endothelial cells [18]. A recent study in a rodent model discovered a protective effect of bilirubin against diabetic nephropathy through inhibition of renal nicotinamide adenine dinucleotide phosphate-(NADPH-) dependent superoxide production and both hyperglycemia and angiotensin-II-induced production of reactive oxygen species [19]. Taken together, the present results suggest that bilirubin might have a protective role in the progression of diabetic nephropathy, particularly in diabetic patients with greater oxidative stress such as CVD. In our study, the standardized coefficient for eGFR was significant in the subgroup with prevalence of CVD, and there were significant interactions between the two groups regarding CVD.
We must be aware of the limitations to the present study. First, due to the cross-sectional study design, the present results are inherently limited in the ability to eliminate causal relationships between serum bilirubin and eGFR among diabetic patients. Second, our definition of eGFR is based on a single assessment of serum creatinine, which may introduce a misclassification bias. Third, estimating GFR listed as the CKD-EPI equation tends to be less accurate in subjects with normal renal function and CKD than GFR when inulin clearance is used, but is more accurate than serum creatinine or eGFR using the Modification of Diet in Renal Disease (MDRD) formula [14]. Fourth, in this study, CKD may have been misclassified with eGFR > 60 mL/min/1.73 m 2 and proteinuria as mildly reduced renal function because renal dysfunction was defined as reduced eGFR irrespective of the presence or absence of proteinuria. Therefore, the demographics and referral source may limit generalizability.

Conclusion
The present study showed that decreased serum bilirubin levels are strongly associated with decreased eGFR among diabetic patients. The underlying mechanism behind this relationship is unclear, but seems to be independent of traditional confounding risk factors such as age, hypertension, and dyslipidemia. For community-dwelling diabetic patients, prospective population-based studies are needed to investigate the mechanisms underlying this association.