- Email: [email protected]
Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China
JDC-07265; No of Pages 6 Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Journal of Diabetes and Its Complications journal homepage: www.jdcjournal.com
Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China Yanfeng Ren a,d,1, Leili Gao b,1, Xiaohui Guo c, Xiaoxu Huo d, Juming Lu e, Jing Li d, Linong Ji b,⁎, Xilin Yang d,⁎⁎ a
Department of Health Statistics, School of Public Health and Management, Weifang Medical University, Shandong, China Department of Endocrinology, Peking University People's Hospital, Beijing, China Department of Endocrinology, Peking University First Hospital, Beijing, China d Department of Epidemiology and Biostatistics, School of Public Health, Tianjin Medical University, Tianjin, China e Department of Endocrinology, Chinese PLA General Hospital, Beijing, China b c
a r t i c l e
i n f o
Article history: Received 28 November 2017 Received in revised form 22 April 2018 Accepted 2 September 2018 Available online xxxx Keywords: Uric acid Bilirubin Type 2 diabetes mellitus Micro-vascular disease Chinese
a b s t r a c t Aims: Serum uric acid (SUA) and bilirubin at high levels had both pro-oxidant and anti-oxidant properties. The present study aimed to examine additive interactions between SUA and total bilirubin (TBIL) for the risk of micro-vascular disease (MVD) in type 2 diabetes mellitus (T2DM). Methods: A cross-sectional survey of 6713 inpatients with T2DM was conducted in 81 tertiary care hospitals in China. MVD was defined as having either prior diabetic retinopathy (DR) or diabetic nephropathy (DN). Binary logistic regression was used to estimate odds ratios of SUA and TBIL for MVD. Additive interaction was measured by three indices, i.e., relative excess risk due to interaction, attributable proportion due to interaction and synergy index. Results: Among 6713 inpatients, 408 (6.08%) suffered from MVD. SUA ≥ 283 μmol/l (i.e., its media) was defined as high SUA, and TBIL b11.5 μmol/l (n = 2290 or 34.11%) was defined as low TBIL. Overall, 621 patients were exposed to co-presence of high SUA and low TBIL. The co-presence of both factors greatly increased the effect sizes from 1.03(95%CI: 0.72–1.46) (high SUA alone) or 0.70(95%CI: 0.48–1.05) (low TBIL alone) to 1.90 (95%CI: 1.26–2.87) for MVD in multivariable analysis. The additive interaction of both factors was significant for MVD in both univariable analysis and multivariable analysis. Conclusions: Co-presence of both high SUA and low TBIL indentified a group of patients at a markedly increased risk of MVD in high-risk Chinese patients with T2DM. © 2018 Elsevier Inc. All rights reserved.
1. Introduction Life style transition and population aging have led to a rapid increase in the prevalence of diabetes worldwide, 1 and more and more people are suffering from its complications. How to reduce the burden of diabetes and its complications has become a health priority in many parts of the world. Diabetic retinopathy (DR) and diabetic nephropathy (DN) are two of the chronic micro-vascular complications of diabetes. DR is a major cause of vision loss in adults2 while DN is the leading cause of life-threatening end-stage renal disease (ESRD). 3 Identification of novel risk factors is certain to contribute to a further reduction in the risk of micro-vascular disease (MVD) in type 2 diabetes mellitus Financial Disclosure: L.J received research grant/s from Novo Nordisk China. ⁎ Correspondence to: L. Ji, Department of Endocrinology, Peking University People's Hospital, Xicheng District, Beijing 100044, China. ⁎⁎ Correspondence to: X. Yang, P.O. Box 154, School of Public Health, Tianjin Medical University, 22 Qixiangtai Road, Heping District, Tianjin 300070, China. E-mail addresses: [email protected] (L. Ji), [email protected] (X. Yang). 1 Equal contribution to the manuscript.
(T2DM). In this regard, serum uric acid (SUA) and bilirubin were recently reported to be associated with MVD in diabetes 4,5 but it remains unknown whether the two risk factors have a synergistic or interactive effect on the risk of MVD in T2DM. Uric acid, a metabolic end-product of purine degradation, possesses pro-oxidant property in human and is also a promoter of oxidative and inflammatory processes.6 The levels of SUA are influenced by a variety of factors, including dietary intake, and its rates of production and renal excretion. Elevated serum uric acid levels may stem from imbalance between production and excretion. Because of lack of consensus in the diagnostic criteria for hyperuricemia, most studies defined hyperuricemia as SUA N7.0 mg/dl in man and 6.0 mg/dl in women. 7,8 Accumulation of monosodium urate crystals in the joint fluid and periarticular tissue may result in gout, one of major consequence of hyperuricemia. Recent studies revealed that hyperuricemia was associated with increased risks of cardiovascular disease, chronic kidney disease, diabetes and its complications.5,9–11 Elevated SUA even in the normal range is associated with increased risks of diabetes and vascular complications. 12,13
https://doi.org/10.1016/j.jdiacomp.2018.09.002 1056-8727/© 2018 Elsevier Inc. All rights reserved.
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
2
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
Bilirubin is an end product of heme metabolism and considered to be a toxic metabolite in jaundice, particularly in neonates. However, recent studies found that bilirubin exhibited a cytoprotectant effect and possessed both antioxidant and anti-inflammatory properties on the vasculature. 14,15 Several cross-sectional and longitudinal studies revealed that bilirubin in human may have beneficial effects on hypertension, atherosclerosis, metabolic syndrome, cardiovascular disease and diabetes.16–20 In patients with diabetes, serum bilirubin levels are not only inversely associated with risk of cardiovascular disease but also associated with risk of MVD 4,21 and severity of these complications. 21 A prospective study demonstrated a U-shaped relationship between bilirubin levels and the risk of coronary heart disease.22 Given to the anti-oxidative and anti-inflammatory properties of SUA and serum bilirubin, and their associations with MVD, it is worthwhile to explore possible interactive effects of high SUA and low TBIL on the risk of MVD in T2DM. The current study used the data from a cross sectional survey of inpatients with T2DM from Chinese 81 top tertiary care hospitals to examine 1) full-range associations of SUA and TBIL with MVD; 2) to test interactive effects of both factors on the risk of MVD in Chinese patients with T2DM. 2. Methods 2.1. Patients Chinese Hospital Association conducted a hospital-based crosssectional survey from May 2013 to August 2013 as a quality improvement effort in the management of patients with T2DM in tertiary care hospitals in China. A total of 81 top tertiary care hospitals from 21 provinces in China participated in the survey. The inclusion criteria were 1) patients with T2DM and admitted to the division of endocrinology; 2) agreed the treatment scheme of basal bolus plus meal time insulin after admission; 3) between the ages of 18 to 80 years. The exclusion criteria were 1) with alanine aminotransferase or aspartate aminotransferase ≥100 U/l; 2) with serum creatinine ≥110 μmol/l in female and ≥125 μmol/l in male or chronic kidney disease; 3) during pregnancy or lactation or both; and 4) unable to communicate in a normal way. The detailed patient inclusion and exclusion criteria were available elsewhere.23 During the 6-month fieldwork period, a total of 6800 inpatients with T2DM were consecutively recruited. Among them, 87 patients were excluded due to missing key variables or meeting exclusion criteria. The remaining 6713 patients were used in the final analyses. The ethics approval was granted by the People's Liberation Army (PLA) General Hospital Clinical Research Ethics Committee and written informed consent was obtained from every patient before data collection. 2.2. Data collection and clinical measurements Postgraduate medical students or research nurses conducted the fieldwork after being trained in a workshop to standardize all the data collection procedures. Patients were asked to fill out a structured questionnaire to collect data on demographics, lifestyle and medical history. Case notes were reviewed to record clinical characteristics and laboratory measurements, and then transformed the data into the questionnaire. The demographic and clinical characteristics included age, gender, duration of diabetes, body height, weight, blood pressure (BP), and self-monitoring of blood glucose (SMBG). Blood pressure was taken in the seated position using standardized sphygmomanometers. Body mass index (BMI) was calculated as the weight in kilograms divided by the square of the height in meters. Measurements of lowdensity lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), glycated hemoglobin (HbA1c), SUA, TBIL, liver function and renal function were performed in the laboratory of local hospitals. Overnight fasting blood (at least 8 h of fasting) was taken and plasma was used in the above measurements. First
morning urine was collected to test both albuminuria and creatinine. Use of drugs at admission and during stay in the hospital was documented, including antidiabetes drugs (oral antidiabetes drugs, insulin and glucagon-like peptide-1 based drugs and combined use of these drugs), antihypertensive drugs (calcium antagonists, renin angiotensin system inhibitors and β- receptor antagonists), and lipid lowering drugs (statins and other lipid lowing drugs). The complications of diabetes were documented in details, including macro- and micro- vascular disease. Macro-vascular disease (CVD) was defined as having any of coronary heart disease (CHD), stroke or peripheral arterial disease (PAD). The detailed definitions of CHD, stroke and PAD were published previously.23 2.3. Definition of micro-vascular disease DN was diagnosed by presence of persistent proteinuria, i.e., urinary albumin (UA) ≥ 30 mg/24 h or urinary albumin excretion rate (UAER) ≥ 20 μg/min for at least three consecutive readings after excluding blood urine, urinary system infection and kidney damage from other causes. At least one ophthalmologist conducted ophthalmoscopy and fluorescein angiography by dilated pupils in the hospital, where patients had ever seen the doctor. DR was defined as background of retinopathy, pre-proliferative, proliferative, and maculopathy. Considered that the mechanism of peripheral neuropathy may be different from DR and DN, 24,25 micro-vascular disease was defined as having either DN or DR in this study. 2.4. Statistical analysis SAS 9.3 (SAS Institute, Inc., Cary, NC, USA) was used in the current analysis. Continuous variables were expressed as medians and interquartiles or means and standard deviations where appropriate. Categorical variables were expressed as numbers and percentages. P-P plot and Q-Q plot were used to determine normal distribution of continuous variables. Categorical variables between patients with MVD and without MVD were compared using Chi-square test or Fisher's exacttest where appropriate. Continuous variables between two groups were compared by two-sample Wilcoxon rank test if normal distribution was rejected or Student t-test otherwise. Binary logistic regression was used to estimate associations of TBIL and SUA with MVD. A structured adjustment scheme was used to adjust for confounding effects of covariables. First, OR(95%CI)s of variables under investigation for MVD were estimated in univariable models. Then, we performed multivariable analysis to obtain adjusted OR(95%CI)s of TBIL and SUA for MVD. Adjusted variables included age, gender, duration of diabetes, BMI, systolic blood pressure, diastolic blood pressure, SMBG, HbA1c, LDL-C, HDL-C, TG, use of drugs (oral antidiabetes drugs, insulin, lipid lowering drugs and antihypertensive drugs), and complications of diabetes(macro-vascular disease, sensory neuropathy, and other complications). We used restricted cubic splines (RCS) in binary logistic regression analyses to check full-range associations of SUA and TBIL with MVD. Four knots (5%, 35%, 65% and 95%) were selected in RCS analysis as suggested by Harrell.26 Cutoff points were chosen based on visual checking of the curves, i.e., where the risk of MVD rapidly increased or decreased after the cutoff points. Then, SUA and TBIL were stratified into binary categorical variables if appropriate. Additional binary logistic regression analyses were performed to ascertain the effect sizes of high versus low SUA (or low versus high TBIL). Additive interactions were tested by three measures, i.e., relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP) and synergy index (S). Significant RERI (95%CI) N0, AP (95% CI) N0, or S (95%CI) N1 indicates significant additive interaction between the two variables. 27 P values b 0.05 from two-sided tests were considered to be statistically significant.
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
3
Table 1 Demographic, clinical and biochemical characteristics of subjects by MVD.
Variable Age, years Male gender Duration of diabetes, years Body mass index, kg/m2 HbA1c, % HbA1c, mmol/mol Systolic blood pressure, mmHg Diastolic blood pressure, mmHg LDL-C, mmol/l HDL-C, mmol/l TG, mmol/l SMBG, yes Serum uric acid, μmol/l ≥283 μmol/l Total bilirubin, μmol/l b11.5 μmol/l Complications Diabetic nephropathy Diabetic retinopathy CHD Stroke PAD Sensory neuropathy Diabetic foot Hypertension Drug use Stains Other lipid lowering drugs Renin-angiotensin system inhibitors Other antihypertensive drugs OADs only GLP-1 based drugs alone or combined with OADs Insulin alone or combined with OADs
MVD (n = 408)
Non MVD (n = 6035)
Median (25th to 75th) or n(%)
Median (25th to 75th) or n(%)
59(52–65) 234(57.35) 7.96(4.89–13.10) 24.12(21.71–26.59) 10.20(9.10–11.20) 88(76–99) 130.0(123.5–140.0) 80(77–90) 3.03(2.49–3.60) 1.20(0.98–1.70) 2.09(1.37–2.70) 153(37.50) 130.0(123.5–140.0) 222(54.41) 11.8(8.7–13.9) 168(41.18)
57(50–64) 3563(56.51) 2.94(0.37–5.90) 23.80(21.82–25.68) 10.20(9.60–11.20) 88(81–99) 130.0(124.0–135.0) 80(75–85) 3.00(2.69–3.70) 1.90(1.43–2.90) 2.30(1.56–3.00) 1962(31.12) 130.0(124.0–135.0) 3139(49.79) 12.0(10.6–12.8) 2122(33.66)
P value
b0.0001⁎ 0.7394⁎ b0.0001⁎ 0.0071⁎ 0.1869† 0.1869† b0.0001⁎ b0.0001⁎
0.0785† b0.0001† b0.0001† 0.0072⁎ 0.1602† 0.0701⁎ 0.9858† 0.0019⁎
220(53.92) 281(68.87) 103(25.25) 30(7.57) 100(24.51) 226(55.39) 17(4.17) 182(44.61)
0(0.00) 0(0.00) 254(4.03) 76(1.21) 116(1.84) 204(3.24) 49(0.78) 321(5.09)
b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎
130(31.86) 15(3.68) 136(33.33) 91(22.30) 169(41.42) 3(0.74) 197(48.28)
214(3.39) 49(0.78) 262(4.16) 114(1.81) 2472(39.21) 24(0.38) 2089(33.13)
b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ 0.3748⁎ 0.2240⁎ b0.0001⁎
MVD, micro-vascular disease; HbA1c, glycated hemoglobin; SMBG, self-monitoring of blood glucose; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; OADs, oral antidiabetes drugs; GLP, glucagon-like peptide. ⁎ P values were derived from Chi-square test or t-test. † P values were derived from Wilcoxon rank test.
Additional sensitivity analysis was performed to check the stability of associations of SUA and TBIL with MVD and their interactive effect on MVD. Because serum levels of uric acid may be affected by alterations in renal excretion, we performed a sensitivity analysis of patients who did not use diuretics, i.e., excluding 30 patients who used diuretics. 3. Results 3.1. Characteristics of the study patients Among 6713 patients with T2DM, female patients accounted for 43.44%. Mean age was 56.38 (standard deviation: 10.55) years with 3.00 (interquartile: 0.41 to 6.05) years of duration of diabetes. A total of 408 (6.08%) patients suffered from MVD (DR: 220 or 3.28%; DR: 281 or 4.19%). Patients with MVD were more likely to have an older age, longer duration of diabetes, higher BPs and BMI, and they were more likely to use SMBG than those without MVD. HDL-C and TG were lower in patients with MVD than in those without MVD. Patients with MVD were more likely to have almost all diabetes complications of macro-vascular disease, sensory neuropathy, and diabetic foot. Patients with MVD were more likely to use statins, renin angiotensin system inhibitors and insulin/oral andtidiabetes drugs (OADS) than those without (Table 1). 3.2. Associations of SUA and TBIL with MVD Both SUA and TBIL were associated with the risk of MVD in nonlinear manner. In multivariable analysis, SUA was positively associated
with the risk of MVD in a roughly linear manner up to 300 μmol/l and then, leveled off from 300 μmol/l onwards, without a clear cutoff effect. On the other hand, the risk of MVD was similar in TBIL levels from 17 μmol/l to 11.5 μmol/l and linearly decreased from 11.5 μmol/l to 9 μmol/l downwards and then leveled off from 9 μmol/l downwards (Fig. 1). If SUA was categorized into a binary variable at its median, 283 μmol/l and TBIL was stratified into a binary variable at 11.5 μmol/l, high SUA was associated with increased risk of MVD in multivariable model (OR: 1.46, 95%CI: 1.11–1.93). Low TBIL was significantly associated with increased risk for MVD in univariable model (OR: 1.38, 95%CI: 1.12–1.69) but not in multivariable model (OR: 1.01, 95%CI: 0.76–1.33) (Table 2).
3.3. Subgroup effects of SUA and TBIL on MVD The ORs of high versus low SUA were highly significant among patients with TBIL b11.5 μmol/l in univariable analysis (OR: 3.94, 95%CI: 2.86–5.44) and in multivariable analysis (2.29, 95%CI: 1.50–3.49). On the other hand, the OR of high versus low SUA was greatly attenuated to non-significance among patients with TBIL ≥11.5 μmol/l in multivariable analysis (1.12, 95%CI: 0.76–1.64). Similarly, the ORs of low versus high TBIL were highly significant among patients with SUA ≥ 283 μmol/l in univariable analysis (3.71, 95%CI: 2.80–4.92) and in multivariable analysis (1.53, 95%CI: 1.02–2.31) but not among patients with SUA b283 μmol/l in multivariable analysis (OR: 0.81, 95% CI: 0.54–1.22) (Table 2).
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
4
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
Fig. 1. Odds ratio curves SUA and TBIL for MVD in Chinese patients with type 2 diabetes. Black curves were derived from univariable model; Red curves were derived from multivariate models adjusted for age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, highdensity lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1 and spline functions of SUA, and TBIL were used to adjust for their non-linear confounding effects. Abbreviations: SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease.
3.4. Interactive effects of SUA and TBIL on MVD
4. Discussion
If patients with SUA b283 μmol/l and TBIL ≥11.5 μmol/l were used as the referent group, presence of SUA ≥283 μmol/l alone or TBIL ≤11.5 μmol/l alone was not associated with increased risk of MVD in multivariable analysis. On the other hand, co-presence of both factors was associated with 1.90-fold risk of MVD in multivariable analysis (95% CI, 1.26–2.87) (Table 3). The measures for testing additive interactions were significant in both univariable analysis (RERI: 2.19, 95%CI: 1.57–2.82; AP: 0.87, 95%CI: 0.76–0.98) and multivariable analysis (RERI: 1.16, 95%CI: 0.48–1.85; AP: 0.61, 95%CI: 0.37–0.86) (Table 4).
In this cross-sectional study of inpatients with T2DM, we found that high SUA, i.e., ≥283 μmol/l and low TBIL, i.e., b11.5 μmol/l, had a synergistic effect to increase the risk of MVD in T2DM but presence of high SUA alone or low TBIL alone was not associated with increased risk of MVD. Many epidemiological studies consistently demonstrated a positive association between SUA levels and risk of MVD in patients with T2DM, 5,12 and uric acid-lowering treatment improved symptoms of MVD in a randomized trial. 28 A cross-sectional study from Taiwan in patients with T2DM demonstrated that a high uric acid concentration was a risk factor for albuminuria (OR: 1.23, 95%CI: 1.02–1.48) and DR (OR: 1.26, 95%CI: 1.08–1.47). 5 A 3-year prospective study reported that patients with SUA in the 3rd (5.9–6.9 mg/dl) and 4th (≥7.0 mg/ dl) quartiles had increased risks of DR worsening than patients with SUA in the 1st quartile (b4.9 mg/dl) (Hazard ratio: 2.57, 95%CI: 1.30 to 5.08; and 3.66, 95%CI: 1.92 to 7.00). 12 In a large longitudinal cohort study of patients with T2DM (n = 62,830), patients with 2nd, 3rd,
3.5. Sensitivity analysis of the relationship of SUA and TBIL with MVD In addition, sensitivity analysis showed that the effect size of SUA and TBIL for MVD was similar in patients who didn't use diuretics. The RERI and AP were all significant (Data no shown).
Table 2 Odds ratios of SUA and TBIL for MVD in type 2 diabetes. Exposures
N (%) of MVDc
OR (95% CI)
P value
All patients Model 1a: SUA ≥283 μmol/l vs. b283 μmol/l Model 2b: SUA ≥283 μmol/l vs. b283 μmol/l Model 1a: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l Model 2b: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l
222(6.61):186(5.55) 222(6.61):186(5.55) 168(7.34):240(5.43) 168(7.34):240(5.43)
1.20(0.98, 1.47) 1.46(1.11, 1.93) 1.38(1.12, 1.69) 1.01(0.76, 1.33)
0.0705 0.0068 0.0020 0.9500
Among patients with TBIL ≥ 11.5 μmol/l Model 1a: SUA ≥283 μmol/l vs. b283 μmol/l Model 2b: SUA ≥283 μmol/l vs. b283 μmol/l
127(4.64): 113(6.71) 127(4.64): 113(6.71)
0.67(0.52, 0.87) 1.12(0.76, 1.64)
0.0032 0.5671
Among patients with TBIL b 11.5 μmol/l Model 1a: SUA ≥283 μmol/l vs. b283 μmol/l Model 2b: SUA ≥283 μmol/l vs. b283 μmol/l
95(15.30): 73(4.37) 95(15.30): 73(4.37)
3.94(2.86, 5.44) 2.29(1.50, 3.49)
b0.0001 0.0001
Among patients with SUA b 283 μmol/l Model 1a: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l Model 2b: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l
73(4.37): 113(6.71) 73(4.37): 113(6.71)
0.63(0.47, 0.86) 0.81(0.54, 1.22)
0.0033 0.3177
Among patients with SUA ≥ 283 μmol/l Model 1a: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l Model 2b: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l
95(15.30): 127(4.64) 95(15.30): 127(4.64)
3.71(2.80, 4.92) 1.53(1.02, 2.31)
b0.0001 0.0422
SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease; N (%), number of cases (% of number at risk); OR, odds ratios; CI, confidence interval. a Univariable model, not adjusted for any other variables. b Multivariable model, age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1,were adjusted in multivariable analysis. c N (%), number of cases (% of number at risk).
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
5
Table 3 Interactive effect of SUA and TBIL on MVD in type 2 diabetes. Exposures
N (%) of MVDc
OR (95% CI)
P value
a
Model 1 SUA b 283 SUA ≥ 283 SUA b 283 SUA ≥ 283 Model 2b SUA b 283 SUA ≥ 283 SUA b 283 SUA ≥ 283
μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL
≥ 11.5 μmol/l ≥ 11.5 μmol/l b 11.5 μmol/l b 11.5 μmol/l
113(6.71) 127(4.64) 73(4.37) 95(15.30)
1 0.67(0.52, 0.87) 0.63(0.47, 0.86) 2.50(1.87, 3.35)
b0.0001 b0.0001 b0.0001
μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL
≥ 11.5 μmol/l ≥ 11.5 μmol/l b 11.5 μmol/l b 11.5 μmol/l
113(6.71) 127(4.64) 73(4.37) 95(15.30)
1 1.03(0.72, 1.46) 0.70(0.48, 1.05) 1.90(1.26, 2.87)
0.8897 0.0819 0.0023
SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease; N (%), number of cases (% of number at risk); OR, odds ratios; CI, confidence interval. a Univariable model, not adjusted for any other variables. b Multivariable model, age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1, were adjusted in multivariable analysis. c N (%), number of cases (% of number at risk).
4th, and 5th quintiles of SUA had markedly increased risks of CKD than those with the lowest quintile of SUA (Hazard ratio: 1.46, 95%CI: 1.14 to 1.8; 1.44, 95%CI: 1.11 to 1.87; 1.95, 95%CI: 1.48 to 2.58 and 2.61, 95%CI: 1.98 to 3.42, respectively. 29 A small double-blinded randomized controlled trial showed that use of low-dose allopurinol reduced severity of proteinuria, possibly via decreasing the serum uric acid level.28 Consistently, our study found that elevated SUA was also associated with increased risk of MVD in T2DM. Many cross-sectional and prospective studies reported an inverse relationship between serum bilirubin and MVD. A study in Japanese patients with T2DM showed that higher TBIL was independently associated with decreased risk of DN (OR: 0.096, 95% CI: 0.034 to 0.273) and DR (OR: 0.242, 95%CI: 0.096 to 0.615). 30 A population-based crosssectional study of 1761 patients with T2DM showed that patients in the 4th quartile (i.e. serum bilirubin N 0.99 mg/dl) were less likely to suffer from DR than patients in the 1st quartile of serum bilirubin (i.e. serum bilirubin b 0.60 mg/dl) (OR: 0.55; 95% CI, 0.33 to 0.91). 31 A meta-analysis showed that there was a significant negative relationship between TBIL and the risk of DR (OR: 0.19, 95%CI: 0.14 to 0.25) in a nonlinear manner.32 In a prospective cohort study, higher baseline bilirubin levels were associated with significantly lower risk of progression from micro-albuminuria to macro-albuminuria.33 A study22 demonstrated a U-shaped relationship between bilirubin levels and the risk of coronary heart disease. Differently, our study found that TBIL was associated with the risk of MVD in an L-shaped manner. The different findings in the two studies do not rule out a possible effect of other antioxidants and micronutrients on MVD in T2DM. Indeed, it is warranted to further investigate whether the other antioxidants and micronutrient factors play a role in vascular complications in T2DM.
Table 4 Measures of additive interaction between SUA and TBIL for the risk of MVD. Measures of interaction Univariable model RERI AP S Multivariable model RERI AP S
Estimated value
95%CI
P value
2.19 0.87 −2.19
1.57–2.82 0.76–0.98 NA
b0.0001 b0.0001 NA
1.16 0.61 −3.35
0.48–1.85 0.37–0.86 NA
0.0010 b0.0001 NA
SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease; CI, confidence interval; RERI, relative excess risk of interaction; AP, attributable proportion; S, synergy index; NA, not available. Univariable model: not adjusted for any other variables. Multivariable model: age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1 were adjusted in multivariable analysis.
To our knowledge, our study is the first reporting a synergistic effect between SUA and TBIL for the risk of MVD in patients with T2DM. In our study, co-presence of high SUA and low TBIL had a two-fold MVD risk of those with co-presence of low SUA and high TBIL. Of interest, we also observed that patients with isolated high SUA or low TBIL were not associated with significantly increased risk of MVD. The exact mechanisms responsible for the associations of high uric acid and low bilirubin with micro-vascular disease are still unclear. Uric acid had the capacity of pro-oxidant and decreasing nitric oxide bioavailability. Increased uric acid may stimulate the renninangiotensin system, damage vascular endothelial cells, and cause proliferative effects on vascular smooth muscle. 34 However, bilirubin has been recognized as a substance with potent antioxidant and antiinflammatory properties. 35 In addition, the modulatory effects of bilirubin on T regulatory cell differentiation 36 further underlined the protective role of bilirubin in the pathogenesis of chronic inflammatory. According to the physiological property and role of SUA and TBIL, it is possible that increased SUA or decreased TBIL can cause microvascular damage. If increased SUA and decreased TBIL co-existed, the damage induced by one factor may be interacted with or aggravated by the other, especially in the micro-environment of oxidative stress and activated inflammatory cascade induced by hyperglycemia. Our findings have clear clinical implications. SUA and TBIL are routinely measured as their usefulness in clinical practice and therefore, they may be used as markers for MVD in patients with high SUA and low TBIL. Co-presence of high SUA and low TBIL may be useful for stratification of MVD risk among patients with T2DM, individually or combined with other risk factors. In addition, these patients were not a small group, as they accounted for 9.25% of the patients in the survey. Given to the importance of MVD and a high proportion of patients with both two factors in patients with T2DM, clinicians need to be aware that patients with T2DM exposed to both high SUA and low TBIL had a markedly increased risk of MVD. It remains to see whether multi-factors intensive interventions achieving the glycemia, lipid and BP control targets in T2DM can reduce the increased risk of MVD associated with co-presence of both two factors. Our study has several limitations. First, our study patients came from top tertiary hospitals, which restricted the extrapolation of current results to the general patients with T2DM in China. Second, our study used a cross sectional study design. The findings of this current study needs to be confirmed by prospective studies and randomized controlled trials if indicated. Third, in our study, MVD events were identified by reviewing medical records, not by universal complication screening. Although all inpatients of tertiary hospitals in China had detailed clinical examinations, some MVD cases may be missed. Fourth, lifestyle and social-economic factors, such as smoking, alcohol drinking habits, education levels, income and history of gout were not collected in the study, it was uncertain whether those factors could partially explain
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
6
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
the observed associations reported in the study. Fifth, patients with abnormal kidney function, i.e., serum creatinine ≥125 μmol/l in male and ≥ 110 μmol/l in female or chronic kidney disease were not invited to participate in this survey and the endpoint of diabetes nephropathy only included mild diabetes nephropathy without declined renal function. So the observed associations in our study may underestimate the risk in the general patient population if those patients with declined renal function were included in the analysis. In summary, we found that co-presence of SUA ≥283 μmol/l and TBIL b11.5 μmol/l was associated with markedly increased risk of MVD in Chinese patients with T2DM. Isolated presence of either of them was not associated with increased risks of MVD. Co-presence of SUA ≥283 μmol/l and TBIL b11.5 μmol/l may be a potential predictor for MVD if our findings are confirmed in patients with T2DM, especially among low risk patients with T2DM. Acknowledgements This study was supported by a research grant from Novo Nordisk China. In addition, we thank fieldworkers involved in the survey for their efforts in data collection and quality control. References 1. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 2004;27:1047-53. 2. Fong DS, Aiello LP, Ferris III FL, Klein R. Diabetic retinopathy. Diabetes Care 2004;27: 2540-53. 3. Hwang SJ, Tsai JC, Chen HC. Epidemiology, impact and preventive care of chronic kidney disease in Taiwan. Nephrology (Carlton) 2010;15:3-9. 4. Chan KH, O'Connell RL, Sullivan DR, Hoffmann LS, Rajamani K, Whiting M, et al. Plasma total bilirubin levels predict amputation events in type 2 diabetes mellitus: the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study. Diabetologia 2013;56:724-36. 5. Liang CC, Lin PC, Lee MY, Chen SC, Shin SJ, Hsiao PJ, et al. Association of serum uric acid concentration with diabetic retinopathy and albuminuria in Taiwanese patients with type 2 diabetes mellitus. Int J Mol Sci 2016;17:1248. 6. Kang DH, Ha SK. Uric acid puzzle: dual role as anti-oxidantand pro-oxidant. Electrolyte Blood Press 2014;12:1-6. 7. Gillett MJ. International expert committee report on the role of the A1c assay in the diagnosis of diabetes. Clin Biochem Rev 2009;30:197-200. 8. Kuo CC, Weaver V, Fadrowski JJ, Lin YS, Guallar E, Navas-Acien A. Arsenic exposure, hyperuricemia, and gout in US adults. Environ Int 2015;76:32-40. 9. Takae K, Nagata M, Hata J, Mukai N, Hirakawa Y, Yoshida D, et al. Serum uric acid as a risk factor for chronic kidney disease in a Japanese community- the Hisayama study. Circ J 2016;80:1857-62. 10. Weiner DE, Tighiouart H, Elsayed EF, Griffith JL, Salem DN, Levey AS. Uric acid and incident kidney disease in the community. J Am Soc Nephrol 2008;19:1204-11. 11. Wu YQ, Li J, Xu YX, Wang YL, Luo YY, Hu DY, et al. Predictive value of serum uric acid on cardiovascular disease and all-cause mortality in urban Chinese patients. Chin Med J (Engl) 2010;123:1387-91. 12. Lee JJ, Yang IH, Kuo HK, Chung MS, Chen YJ, Chen CH, et al. Serum uric acid concentration is associated with worsening in severity of diabetic retinopathy among type 2 diabetic patients in Taiwan–a 3-year prospective study. Diabetes Res Clin Pract 2014;106:366-72. 13. Shani M, Vinker S, Dinour D, Leiba M, Twig G, Holtzman EJ, et al. High normal uric acid levels are associated with an increased risk of diabetes in lean, normoglycemic healthy women. J Clin Endocrinol Metab 2016;101:3772-8.
14. Baranano DE, Rao M, Ferris CD, Snyder SH. Biliverdin reductase: a major physiologic cytoprotectant. Proc Natl Acad Sci U S A 2002;99:16093-8. 15. Dekker D, Dorresteijn MJ, Pijnenburg M, Heemskerk S, Rasing-Hoogveld A, Burger DM, et al. The bilirubin-increasing drug atazanavir improves endothelial function in patients with type 2 diabetes mellitus. Arterioscler Thromb Vasc Biol 2011;31: 458-63. 16. Chin HJ, Song YR, Kim HS, Park M, Yoon HJ, Na KY, et al. The bilirubin level is negatively correlated with the incidence of hypertension in normotensive Korean population. J Korean Med Sci 2009;24:S50-6. [Suppl.]. 17. Hwang HJ, Kim SH. Inverse relationship between fasting direct bilirubin and metabolic syndrome in Korean adults. Clin Chim Acta 2010;411:1496-501. 18. Jung CH, Lee MJ, Kang YM, Hwang JY, Jang JE, Leem J, et al. Higher serum bilirubin level as a protective factor for the development of diabetes in healthy Korean men: a 4 year retrospective longitudinal study. Metabolism 2014;63:87-93. 19. Kang SJ, Kim D, Park HE, Chung GE, Choi SH, Choi SY, et al. Elevated serum bilirubin levels are inversely associated with coronary artery atherosclerosis. Atherosclerosis 2013;230:242-8. 20. Kim ES, Mo EY, Moon SD, Han JH. Inverse association between serum bilirubin levels and arterial stiffness in Korean women with type 2 diabetes. PLoS One 2014;9, e109251. 21. Sekioka R, Tanaka M, Nishimura T, Itoh H. Serum total bilirubin concentration is negatively associated with increasing severity of retinopathy in patients with type 2 diabetes mellitus. J Diabetes Complications 2015;29:218-21. 22. Troughton JA, Woodside JV, Young IS, Arveiler D, Amouyel P, Ferrieres J, et al. Bilirubin and coronary heart disease risk in the Prospective Epidemiological Study of Myocardial Infarction (PRIME). Eur J Cardiovasc Prev Rehabil 2007;14:79-84. 23. Ren Y, Jin N, Hong T, Mu Y, Guo L, Ji Q, et al. Interactive effect of serum uric acid and total bilirubin for cardiovascular disease in Chinese patients with type 2 diabetes. Sci Rep 2016;6:36437. 24. Chandy A, Pawar B, John M, Isaac R. Association between diabetic nephropathy and other diabetic microvascular and macrovascular complications. Saudi J Kidney Dis Transpl 2008;19:924-8. 25. Pradeepa R, Anjana RM, Unnikrishnan R, Ganesan A, Mohan V, Rema M. Risk factors for microvascular complications of diabetes among south Indian subjects with type 2 diabetes–the Chennai Urban Rural Epidemiology Study (CURES) Eye Study-5. Diabetes Technol Ther 2010;12:755-61. 26. Harrell F. Regression Modelling Strategies with Applications to Linear Models, Logistic Regression, and Survival Analysis. New York: Spinger-Varlag New York, Inc.. 2001. 27. Yang X, So WY, Ma RC, Kong AP, Lee HM, Yu LW, et al. Low HDL cholesterol, metformin use, and cancer risk in type 2 diabetes: the Hong Kong diabetes registry. Diabetes Care 2011;34:375-80. 28. Momeni A, Shahidi S, Seirafian S, Taheri S, Kheiri S. Effect of allopurinol in decreasing proteinuria in type 2 diabetic patients. Iran J Kidney Dis 2010;4:128-32. 29. De Cosmo S, Viazzi F, Pacilli A, Giorda C, Ceriello A, Gentile S, et al. Serum uric acid and risk of CKD in type 2 diabetes. Clin J Am Soc Nephrol 2015;10:1921-9. 30. Hamamoto S, Kaneto H, Kamei S, Shimoda M, Tawaramoto K, Kanda-Kimura Y, et al. Low bilirubin levels are an independent risk factor for diabetic retinopathy and nephropathy in Japanese patients with type 2 diabetes. Diabetes Metab 2015;41: 429-31. 31. Najam SS, Sun J, Zhang J, Xu M, Lu J, Sun K, et al. Serum total bilirubin levels and prevalence of diabetic retinopathy in a Chinese population. J Diabetes 2014;6:221-7. 32. Zhu B, Wu X, Ning K, Jiang F, Zhang L. The negative relationship between bilirubin level and diabetic retinopathy: a meta-analysis. PLoS One 2016;11, e0161649. 33. Toya K, Babazono T, Hanai K, Uchigata Y. Association of serum bilirubin levels with development and progression of albuminuria, and decline in estimated glomerular filtration rate in patients with type 2 diabetes mellitus. J Diabetes Investig 2014;5: 228-35. 34. Mene P, Punzo G. Uric acid: bystander or culprit in hypertension and progressive renal disease? J Hypertens 2008;26:2085-92. 35. Vitek L. The role of bilirubin in diabetes, metabolic syndrome, and cardiovascular diseases. Front Pharmacol 2012;3:55. 36. Rocuts F, Zhang X, Yan J, Yue Y, Thomas M, Bach FH, et al. Bilirubin promotes de novo generation of T regulatory cells. Cell Transplant 2010;19:443-51.
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
Contents lists available at ScienceDirect
Journal of Diabetes and Its Complications journal homepage: www.jdcjournal.com
Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China Yanfeng Ren a,d,1, Leili Gao b,1, Xiaohui Guo c, Xiaoxu Huo d, Juming Lu e, Jing Li d, Linong Ji b,⁎, Xilin Yang d,⁎⁎ a
Department of Health Statistics, School of Public Health and Management, Weifang Medical University, Shandong, China Department of Endocrinology, Peking University People's Hospital, Beijing, China Department of Endocrinology, Peking University First Hospital, Beijing, China d Department of Epidemiology and Biostatistics, School of Public Health, Tianjin Medical University, Tianjin, China e Department of Endocrinology, Chinese PLA General Hospital, Beijing, China b c
a r t i c l e
i n f o
Article history: Received 28 November 2017 Received in revised form 22 April 2018 Accepted 2 September 2018 Available online xxxx Keywords: Uric acid Bilirubin Type 2 diabetes mellitus Micro-vascular disease Chinese
a b s t r a c t Aims: Serum uric acid (SUA) and bilirubin at high levels had both pro-oxidant and anti-oxidant properties. The present study aimed to examine additive interactions between SUA and total bilirubin (TBIL) for the risk of micro-vascular disease (MVD) in type 2 diabetes mellitus (T2DM). Methods: A cross-sectional survey of 6713 inpatients with T2DM was conducted in 81 tertiary care hospitals in China. MVD was defined as having either prior diabetic retinopathy (DR) or diabetic nephropathy (DN). Binary logistic regression was used to estimate odds ratios of SUA and TBIL for MVD. Additive interaction was measured by three indices, i.e., relative excess risk due to interaction, attributable proportion due to interaction and synergy index. Results: Among 6713 inpatients, 408 (6.08%) suffered from MVD. SUA ≥ 283 μmol/l (i.e., its media) was defined as high SUA, and TBIL b11.5 μmol/l (n = 2290 or 34.11%) was defined as low TBIL. Overall, 621 patients were exposed to co-presence of high SUA and low TBIL. The co-presence of both factors greatly increased the effect sizes from 1.03(95%CI: 0.72–1.46) (high SUA alone) or 0.70(95%CI: 0.48–1.05) (low TBIL alone) to 1.90 (95%CI: 1.26–2.87) for MVD in multivariable analysis. The additive interaction of both factors was significant for MVD in both univariable analysis and multivariable analysis. Conclusions: Co-presence of both high SUA and low TBIL indentified a group of patients at a markedly increased risk of MVD in high-risk Chinese patients with T2DM. © 2018 Elsevier Inc. All rights reserved.
1. Introduction Life style transition and population aging have led to a rapid increase in the prevalence of diabetes worldwide, 1 and more and more people are suffering from its complications. How to reduce the burden of diabetes and its complications has become a health priority in many parts of the world. Diabetic retinopathy (DR) and diabetic nephropathy (DN) are two of the chronic micro-vascular complications of diabetes. DR is a major cause of vision loss in adults2 while DN is the leading cause of life-threatening end-stage renal disease (ESRD). 3 Identification of novel risk factors is certain to contribute to a further reduction in the risk of micro-vascular disease (MVD) in type 2 diabetes mellitus Financial Disclosure: L.J received research grant/s from Novo Nordisk China. ⁎ Correspondence to: L. Ji, Department of Endocrinology, Peking University People's Hospital, Xicheng District, Beijing 100044, China. ⁎⁎ Correspondence to: X. Yang, P.O. Box 154, School of Public Health, Tianjin Medical University, 22 Qixiangtai Road, Heping District, Tianjin 300070, China. E-mail addresses: [email protected] (L. Ji), [email protected] (X. Yang). 1 Equal contribution to the manuscript.
(T2DM). In this regard, serum uric acid (SUA) and bilirubin were recently reported to be associated with MVD in diabetes 4,5 but it remains unknown whether the two risk factors have a synergistic or interactive effect on the risk of MVD in T2DM. Uric acid, a metabolic end-product of purine degradation, possesses pro-oxidant property in human and is also a promoter of oxidative and inflammatory processes.6 The levels of SUA are influenced by a variety of factors, including dietary intake, and its rates of production and renal excretion. Elevated serum uric acid levels may stem from imbalance between production and excretion. Because of lack of consensus in the diagnostic criteria for hyperuricemia, most studies defined hyperuricemia as SUA N7.0 mg/dl in man and 6.0 mg/dl in women. 7,8 Accumulation of monosodium urate crystals in the joint fluid and periarticular tissue may result in gout, one of major consequence of hyperuricemia. Recent studies revealed that hyperuricemia was associated with increased risks of cardiovascular disease, chronic kidney disease, diabetes and its complications.5,9–11 Elevated SUA even in the normal range is associated with increased risks of diabetes and vascular complications. 12,13
https://doi.org/10.1016/j.jdiacomp.2018.09.002 1056-8727/© 2018 Elsevier Inc. All rights reserved.
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
2
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
Bilirubin is an end product of heme metabolism and considered to be a toxic metabolite in jaundice, particularly in neonates. However, recent studies found that bilirubin exhibited a cytoprotectant effect and possessed both antioxidant and anti-inflammatory properties on the vasculature. 14,15 Several cross-sectional and longitudinal studies revealed that bilirubin in human may have beneficial effects on hypertension, atherosclerosis, metabolic syndrome, cardiovascular disease and diabetes.16–20 In patients with diabetes, serum bilirubin levels are not only inversely associated with risk of cardiovascular disease but also associated with risk of MVD 4,21 and severity of these complications. 21 A prospective study demonstrated a U-shaped relationship between bilirubin levels and the risk of coronary heart disease.22 Given to the anti-oxidative and anti-inflammatory properties of SUA and serum bilirubin, and their associations with MVD, it is worthwhile to explore possible interactive effects of high SUA and low TBIL on the risk of MVD in T2DM. The current study used the data from a cross sectional survey of inpatients with T2DM from Chinese 81 top tertiary care hospitals to examine 1) full-range associations of SUA and TBIL with MVD; 2) to test interactive effects of both factors on the risk of MVD in Chinese patients with T2DM. 2. Methods 2.1. Patients Chinese Hospital Association conducted a hospital-based crosssectional survey from May 2013 to August 2013 as a quality improvement effort in the management of patients with T2DM in tertiary care hospitals in China. A total of 81 top tertiary care hospitals from 21 provinces in China participated in the survey. The inclusion criteria were 1) patients with T2DM and admitted to the division of endocrinology; 2) agreed the treatment scheme of basal bolus plus meal time insulin after admission; 3) between the ages of 18 to 80 years. The exclusion criteria were 1) with alanine aminotransferase or aspartate aminotransferase ≥100 U/l; 2) with serum creatinine ≥110 μmol/l in female and ≥125 μmol/l in male or chronic kidney disease; 3) during pregnancy or lactation or both; and 4) unable to communicate in a normal way. The detailed patient inclusion and exclusion criteria were available elsewhere.23 During the 6-month fieldwork period, a total of 6800 inpatients with T2DM were consecutively recruited. Among them, 87 patients were excluded due to missing key variables or meeting exclusion criteria. The remaining 6713 patients were used in the final analyses. The ethics approval was granted by the People's Liberation Army (PLA) General Hospital Clinical Research Ethics Committee and written informed consent was obtained from every patient before data collection. 2.2. Data collection and clinical measurements Postgraduate medical students or research nurses conducted the fieldwork after being trained in a workshop to standardize all the data collection procedures. Patients were asked to fill out a structured questionnaire to collect data on demographics, lifestyle and medical history. Case notes were reviewed to record clinical characteristics and laboratory measurements, and then transformed the data into the questionnaire. The demographic and clinical characteristics included age, gender, duration of diabetes, body height, weight, blood pressure (BP), and self-monitoring of blood glucose (SMBG). Blood pressure was taken in the seated position using standardized sphygmomanometers. Body mass index (BMI) was calculated as the weight in kilograms divided by the square of the height in meters. Measurements of lowdensity lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), glycated hemoglobin (HbA1c), SUA, TBIL, liver function and renal function were performed in the laboratory of local hospitals. Overnight fasting blood (at least 8 h of fasting) was taken and plasma was used in the above measurements. First
morning urine was collected to test both albuminuria and creatinine. Use of drugs at admission and during stay in the hospital was documented, including antidiabetes drugs (oral antidiabetes drugs, insulin and glucagon-like peptide-1 based drugs and combined use of these drugs), antihypertensive drugs (calcium antagonists, renin angiotensin system inhibitors and β- receptor antagonists), and lipid lowering drugs (statins and other lipid lowing drugs). The complications of diabetes were documented in details, including macro- and micro- vascular disease. Macro-vascular disease (CVD) was defined as having any of coronary heart disease (CHD), stroke or peripheral arterial disease (PAD). The detailed definitions of CHD, stroke and PAD were published previously.23 2.3. Definition of micro-vascular disease DN was diagnosed by presence of persistent proteinuria, i.e., urinary albumin (UA) ≥ 30 mg/24 h or urinary albumin excretion rate (UAER) ≥ 20 μg/min for at least three consecutive readings after excluding blood urine, urinary system infection and kidney damage from other causes. At least one ophthalmologist conducted ophthalmoscopy and fluorescein angiography by dilated pupils in the hospital, where patients had ever seen the doctor. DR was defined as background of retinopathy, pre-proliferative, proliferative, and maculopathy. Considered that the mechanism of peripheral neuropathy may be different from DR and DN, 24,25 micro-vascular disease was defined as having either DN or DR in this study. 2.4. Statistical analysis SAS 9.3 (SAS Institute, Inc., Cary, NC, USA) was used in the current analysis. Continuous variables were expressed as medians and interquartiles or means and standard deviations where appropriate. Categorical variables were expressed as numbers and percentages. P-P plot and Q-Q plot were used to determine normal distribution of continuous variables. Categorical variables between patients with MVD and without MVD were compared using Chi-square test or Fisher's exacttest where appropriate. Continuous variables between two groups were compared by two-sample Wilcoxon rank test if normal distribution was rejected or Student t-test otherwise. Binary logistic regression was used to estimate associations of TBIL and SUA with MVD. A structured adjustment scheme was used to adjust for confounding effects of covariables. First, OR(95%CI)s of variables under investigation for MVD were estimated in univariable models. Then, we performed multivariable analysis to obtain adjusted OR(95%CI)s of TBIL and SUA for MVD. Adjusted variables included age, gender, duration of diabetes, BMI, systolic blood pressure, diastolic blood pressure, SMBG, HbA1c, LDL-C, HDL-C, TG, use of drugs (oral antidiabetes drugs, insulin, lipid lowering drugs and antihypertensive drugs), and complications of diabetes(macro-vascular disease, sensory neuropathy, and other complications). We used restricted cubic splines (RCS) in binary logistic regression analyses to check full-range associations of SUA and TBIL with MVD. Four knots (5%, 35%, 65% and 95%) were selected in RCS analysis as suggested by Harrell.26 Cutoff points were chosen based on visual checking of the curves, i.e., where the risk of MVD rapidly increased or decreased after the cutoff points. Then, SUA and TBIL were stratified into binary categorical variables if appropriate. Additional binary logistic regression analyses were performed to ascertain the effect sizes of high versus low SUA (or low versus high TBIL). Additive interactions were tested by three measures, i.e., relative excess risk due to interaction (RERI), attributable proportion due to interaction (AP) and synergy index (S). Significant RERI (95%CI) N0, AP (95% CI) N0, or S (95%CI) N1 indicates significant additive interaction between the two variables. 27 P values b 0.05 from two-sided tests were considered to be statistically significant.
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
3
Table 1 Demographic, clinical and biochemical characteristics of subjects by MVD.
Variable Age, years Male gender Duration of diabetes, years Body mass index, kg/m2 HbA1c, % HbA1c, mmol/mol Systolic blood pressure, mmHg Diastolic blood pressure, mmHg LDL-C, mmol/l HDL-C, mmol/l TG, mmol/l SMBG, yes Serum uric acid, μmol/l ≥283 μmol/l Total bilirubin, μmol/l b11.5 μmol/l Complications Diabetic nephropathy Diabetic retinopathy CHD Stroke PAD Sensory neuropathy Diabetic foot Hypertension Drug use Stains Other lipid lowering drugs Renin-angiotensin system inhibitors Other antihypertensive drugs OADs only GLP-1 based drugs alone or combined with OADs Insulin alone or combined with OADs
MVD (n = 408)
Non MVD (n = 6035)
Median (25th to 75th) or n(%)
Median (25th to 75th) or n(%)
59(52–65) 234(57.35) 7.96(4.89–13.10) 24.12(21.71–26.59) 10.20(9.10–11.20) 88(76–99) 130.0(123.5–140.0) 80(77–90) 3.03(2.49–3.60) 1.20(0.98–1.70) 2.09(1.37–2.70) 153(37.50) 130.0(123.5–140.0) 222(54.41) 11.8(8.7–13.9) 168(41.18)
57(50–64) 3563(56.51) 2.94(0.37–5.90) 23.80(21.82–25.68) 10.20(9.60–11.20) 88(81–99) 130.0(124.0–135.0) 80(75–85) 3.00(2.69–3.70) 1.90(1.43–2.90) 2.30(1.56–3.00) 1962(31.12) 130.0(124.0–135.0) 3139(49.79) 12.0(10.6–12.8) 2122(33.66)
P value
b0.0001⁎ 0.7394⁎ b0.0001⁎ 0.0071⁎ 0.1869† 0.1869† b0.0001⁎ b0.0001⁎
0.0785† b0.0001† b0.0001† 0.0072⁎ 0.1602† 0.0701⁎ 0.9858† 0.0019⁎
220(53.92) 281(68.87) 103(25.25) 30(7.57) 100(24.51) 226(55.39) 17(4.17) 182(44.61)
0(0.00) 0(0.00) 254(4.03) 76(1.21) 116(1.84) 204(3.24) 49(0.78) 321(5.09)
b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎
130(31.86) 15(3.68) 136(33.33) 91(22.30) 169(41.42) 3(0.74) 197(48.28)
214(3.39) 49(0.78) 262(4.16) 114(1.81) 2472(39.21) 24(0.38) 2089(33.13)
b0.0001⁎ b0.0001⁎ b0.0001⁎ b0.0001⁎ 0.3748⁎ 0.2240⁎ b0.0001⁎
MVD, micro-vascular disease; HbA1c, glycated hemoglobin; SMBG, self-monitoring of blood glucose; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglyceride; OADs, oral antidiabetes drugs; GLP, glucagon-like peptide. ⁎ P values were derived from Chi-square test or t-test. † P values were derived from Wilcoxon rank test.
Additional sensitivity analysis was performed to check the stability of associations of SUA and TBIL with MVD and their interactive effect on MVD. Because serum levels of uric acid may be affected by alterations in renal excretion, we performed a sensitivity analysis of patients who did not use diuretics, i.e., excluding 30 patients who used diuretics. 3. Results 3.1. Characteristics of the study patients Among 6713 patients with T2DM, female patients accounted for 43.44%. Mean age was 56.38 (standard deviation: 10.55) years with 3.00 (interquartile: 0.41 to 6.05) years of duration of diabetes. A total of 408 (6.08%) patients suffered from MVD (DR: 220 or 3.28%; DR: 281 or 4.19%). Patients with MVD were more likely to have an older age, longer duration of diabetes, higher BPs and BMI, and they were more likely to use SMBG than those without MVD. HDL-C and TG were lower in patients with MVD than in those without MVD. Patients with MVD were more likely to have almost all diabetes complications of macro-vascular disease, sensory neuropathy, and diabetic foot. Patients with MVD were more likely to use statins, renin angiotensin system inhibitors and insulin/oral andtidiabetes drugs (OADS) than those without (Table 1). 3.2. Associations of SUA and TBIL with MVD Both SUA and TBIL were associated with the risk of MVD in nonlinear manner. In multivariable analysis, SUA was positively associated
with the risk of MVD in a roughly linear manner up to 300 μmol/l and then, leveled off from 300 μmol/l onwards, without a clear cutoff effect. On the other hand, the risk of MVD was similar in TBIL levels from 17 μmol/l to 11.5 μmol/l and linearly decreased from 11.5 μmol/l to 9 μmol/l downwards and then leveled off from 9 μmol/l downwards (Fig. 1). If SUA was categorized into a binary variable at its median, 283 μmol/l and TBIL was stratified into a binary variable at 11.5 μmol/l, high SUA was associated with increased risk of MVD in multivariable model (OR: 1.46, 95%CI: 1.11–1.93). Low TBIL was significantly associated with increased risk for MVD in univariable model (OR: 1.38, 95%CI: 1.12–1.69) but not in multivariable model (OR: 1.01, 95%CI: 0.76–1.33) (Table 2).
3.3. Subgroup effects of SUA and TBIL on MVD The ORs of high versus low SUA were highly significant among patients with TBIL b11.5 μmol/l in univariable analysis (OR: 3.94, 95%CI: 2.86–5.44) and in multivariable analysis (2.29, 95%CI: 1.50–3.49). On the other hand, the OR of high versus low SUA was greatly attenuated to non-significance among patients with TBIL ≥11.5 μmol/l in multivariable analysis (1.12, 95%CI: 0.76–1.64). Similarly, the ORs of low versus high TBIL were highly significant among patients with SUA ≥ 283 μmol/l in univariable analysis (3.71, 95%CI: 2.80–4.92) and in multivariable analysis (1.53, 95%CI: 1.02–2.31) but not among patients with SUA b283 μmol/l in multivariable analysis (OR: 0.81, 95% CI: 0.54–1.22) (Table 2).
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
4
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
Fig. 1. Odds ratio curves SUA and TBIL for MVD in Chinese patients with type 2 diabetes. Black curves were derived from univariable model; Red curves were derived from multivariate models adjusted for age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, highdensity lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1 and spline functions of SUA, and TBIL were used to adjust for their non-linear confounding effects. Abbreviations: SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease.
3.4. Interactive effects of SUA and TBIL on MVD
4. Discussion
If patients with SUA b283 μmol/l and TBIL ≥11.5 μmol/l were used as the referent group, presence of SUA ≥283 μmol/l alone or TBIL ≤11.5 μmol/l alone was not associated with increased risk of MVD in multivariable analysis. On the other hand, co-presence of both factors was associated with 1.90-fold risk of MVD in multivariable analysis (95% CI, 1.26–2.87) (Table 3). The measures for testing additive interactions were significant in both univariable analysis (RERI: 2.19, 95%CI: 1.57–2.82; AP: 0.87, 95%CI: 0.76–0.98) and multivariable analysis (RERI: 1.16, 95%CI: 0.48–1.85; AP: 0.61, 95%CI: 0.37–0.86) (Table 4).
In this cross-sectional study of inpatients with T2DM, we found that high SUA, i.e., ≥283 μmol/l and low TBIL, i.e., b11.5 μmol/l, had a synergistic effect to increase the risk of MVD in T2DM but presence of high SUA alone or low TBIL alone was not associated with increased risk of MVD. Many epidemiological studies consistently demonstrated a positive association between SUA levels and risk of MVD in patients with T2DM, 5,12 and uric acid-lowering treatment improved symptoms of MVD in a randomized trial. 28 A cross-sectional study from Taiwan in patients with T2DM demonstrated that a high uric acid concentration was a risk factor for albuminuria (OR: 1.23, 95%CI: 1.02–1.48) and DR (OR: 1.26, 95%CI: 1.08–1.47). 5 A 3-year prospective study reported that patients with SUA in the 3rd (5.9–6.9 mg/dl) and 4th (≥7.0 mg/ dl) quartiles had increased risks of DR worsening than patients with SUA in the 1st quartile (b4.9 mg/dl) (Hazard ratio: 2.57, 95%CI: 1.30 to 5.08; and 3.66, 95%CI: 1.92 to 7.00). 12 In a large longitudinal cohort study of patients with T2DM (n = 62,830), patients with 2nd, 3rd,
3.5. Sensitivity analysis of the relationship of SUA and TBIL with MVD In addition, sensitivity analysis showed that the effect size of SUA and TBIL for MVD was similar in patients who didn't use diuretics. The RERI and AP were all significant (Data no shown).
Table 2 Odds ratios of SUA and TBIL for MVD in type 2 diabetes. Exposures
N (%) of MVDc
OR (95% CI)
P value
All patients Model 1a: SUA ≥283 μmol/l vs. b283 μmol/l Model 2b: SUA ≥283 μmol/l vs. b283 μmol/l Model 1a: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l Model 2b: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l
222(6.61):186(5.55) 222(6.61):186(5.55) 168(7.34):240(5.43) 168(7.34):240(5.43)
1.20(0.98, 1.47) 1.46(1.11, 1.93) 1.38(1.12, 1.69) 1.01(0.76, 1.33)
0.0705 0.0068 0.0020 0.9500
Among patients with TBIL ≥ 11.5 μmol/l Model 1a: SUA ≥283 μmol/l vs. b283 μmol/l Model 2b: SUA ≥283 μmol/l vs. b283 μmol/l
127(4.64): 113(6.71) 127(4.64): 113(6.71)
0.67(0.52, 0.87) 1.12(0.76, 1.64)
0.0032 0.5671
Among patients with TBIL b 11.5 μmol/l Model 1a: SUA ≥283 μmol/l vs. b283 μmol/l Model 2b: SUA ≥283 μmol/l vs. b283 μmol/l
95(15.30): 73(4.37) 95(15.30): 73(4.37)
3.94(2.86, 5.44) 2.29(1.50, 3.49)
b0.0001 0.0001
Among patients with SUA b 283 μmol/l Model 1a: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l Model 2b: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l
73(4.37): 113(6.71) 73(4.37): 113(6.71)
0.63(0.47, 0.86) 0.81(0.54, 1.22)
0.0033 0.3177
Among patients with SUA ≥ 283 μmol/l Model 1a: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l Model 2b: TBIL b11.5 μmol/l vs. ≥11.5 μmol/l
95(15.30): 127(4.64) 95(15.30): 127(4.64)
3.71(2.80, 4.92) 1.53(1.02, 2.31)
b0.0001 0.0422
SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease; N (%), number of cases (% of number at risk); OR, odds ratios; CI, confidence interval. a Univariable model, not adjusted for any other variables. b Multivariable model, age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1,were adjusted in multivariable analysis. c N (%), number of cases (% of number at risk).
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
5
Table 3 Interactive effect of SUA and TBIL on MVD in type 2 diabetes. Exposures
N (%) of MVDc
OR (95% CI)
P value
a
Model 1 SUA b 283 SUA ≥ 283 SUA b 283 SUA ≥ 283 Model 2b SUA b 283 SUA ≥ 283 SUA b 283 SUA ≥ 283
μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL
≥ 11.5 μmol/l ≥ 11.5 μmol/l b 11.5 μmol/l b 11.5 μmol/l
113(6.71) 127(4.64) 73(4.37) 95(15.30)
1 0.67(0.52, 0.87) 0.63(0.47, 0.86) 2.50(1.87, 3.35)
b0.0001 b0.0001 b0.0001
μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL μmol/l and TBIL
≥ 11.5 μmol/l ≥ 11.5 μmol/l b 11.5 μmol/l b 11.5 μmol/l
113(6.71) 127(4.64) 73(4.37) 95(15.30)
1 1.03(0.72, 1.46) 0.70(0.48, 1.05) 1.90(1.26, 2.87)
0.8897 0.0819 0.0023
SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease; N (%), number of cases (% of number at risk); OR, odds ratios; CI, confidence interval. a Univariable model, not adjusted for any other variables. b Multivariable model, age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1, were adjusted in multivariable analysis. c N (%), number of cases (% of number at risk).
4th, and 5th quintiles of SUA had markedly increased risks of CKD than those with the lowest quintile of SUA (Hazard ratio: 1.46, 95%CI: 1.14 to 1.8; 1.44, 95%CI: 1.11 to 1.87; 1.95, 95%CI: 1.48 to 2.58 and 2.61, 95%CI: 1.98 to 3.42, respectively. 29 A small double-blinded randomized controlled trial showed that use of low-dose allopurinol reduced severity of proteinuria, possibly via decreasing the serum uric acid level.28 Consistently, our study found that elevated SUA was also associated with increased risk of MVD in T2DM. Many cross-sectional and prospective studies reported an inverse relationship between serum bilirubin and MVD. A study in Japanese patients with T2DM showed that higher TBIL was independently associated with decreased risk of DN (OR: 0.096, 95% CI: 0.034 to 0.273) and DR (OR: 0.242, 95%CI: 0.096 to 0.615). 30 A population-based crosssectional study of 1761 patients with T2DM showed that patients in the 4th quartile (i.e. serum bilirubin N 0.99 mg/dl) were less likely to suffer from DR than patients in the 1st quartile of serum bilirubin (i.e. serum bilirubin b 0.60 mg/dl) (OR: 0.55; 95% CI, 0.33 to 0.91). 31 A meta-analysis showed that there was a significant negative relationship between TBIL and the risk of DR (OR: 0.19, 95%CI: 0.14 to 0.25) in a nonlinear manner.32 In a prospective cohort study, higher baseline bilirubin levels were associated with significantly lower risk of progression from micro-albuminuria to macro-albuminuria.33 A study22 demonstrated a U-shaped relationship between bilirubin levels and the risk of coronary heart disease. Differently, our study found that TBIL was associated with the risk of MVD in an L-shaped manner. The different findings in the two studies do not rule out a possible effect of other antioxidants and micronutrients on MVD in T2DM. Indeed, it is warranted to further investigate whether the other antioxidants and micronutrient factors play a role in vascular complications in T2DM.
Table 4 Measures of additive interaction between SUA and TBIL for the risk of MVD. Measures of interaction Univariable model RERI AP S Multivariable model RERI AP S
Estimated value
95%CI
P value
2.19 0.87 −2.19
1.57–2.82 0.76–0.98 NA
b0.0001 b0.0001 NA
1.16 0.61 −3.35
0.48–1.85 0.37–0.86 NA
0.0010 b0.0001 NA
SUA, serum uric acid; TBIL, total bilirubin; MVD, micro-vascular disease; CI, confidence interval; RERI, relative excess risk of interaction; AP, attributable proportion; S, synergy index; NA, not available. Univariable model: not adjusted for any other variables. Multivariable model: age, duration of diabetes, gender, body mass index, systolic blood pressure, diastolic blood pressure, glycated hemoglobin, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, self-monitoring, drug use and complications as listed in Table 1 were adjusted in multivariable analysis.
To our knowledge, our study is the first reporting a synergistic effect between SUA and TBIL for the risk of MVD in patients with T2DM. In our study, co-presence of high SUA and low TBIL had a two-fold MVD risk of those with co-presence of low SUA and high TBIL. Of interest, we also observed that patients with isolated high SUA or low TBIL were not associated with significantly increased risk of MVD. The exact mechanisms responsible for the associations of high uric acid and low bilirubin with micro-vascular disease are still unclear. Uric acid had the capacity of pro-oxidant and decreasing nitric oxide bioavailability. Increased uric acid may stimulate the renninangiotensin system, damage vascular endothelial cells, and cause proliferative effects on vascular smooth muscle. 34 However, bilirubin has been recognized as a substance with potent antioxidant and antiinflammatory properties. 35 In addition, the modulatory effects of bilirubin on T regulatory cell differentiation 36 further underlined the protective role of bilirubin in the pathogenesis of chronic inflammatory. According to the physiological property and role of SUA and TBIL, it is possible that increased SUA or decreased TBIL can cause microvascular damage. If increased SUA and decreased TBIL co-existed, the damage induced by one factor may be interacted with or aggravated by the other, especially in the micro-environment of oxidative stress and activated inflammatory cascade induced by hyperglycemia. Our findings have clear clinical implications. SUA and TBIL are routinely measured as their usefulness in clinical practice and therefore, they may be used as markers for MVD in patients with high SUA and low TBIL. Co-presence of high SUA and low TBIL may be useful for stratification of MVD risk among patients with T2DM, individually or combined with other risk factors. In addition, these patients were not a small group, as they accounted for 9.25% of the patients in the survey. Given to the importance of MVD and a high proportion of patients with both two factors in patients with T2DM, clinicians need to be aware that patients with T2DM exposed to both high SUA and low TBIL had a markedly increased risk of MVD. It remains to see whether multi-factors intensive interventions achieving the glycemia, lipid and BP control targets in T2DM can reduce the increased risk of MVD associated with co-presence of both two factors. Our study has several limitations. First, our study patients came from top tertiary hospitals, which restricted the extrapolation of current results to the general patients with T2DM in China. Second, our study used a cross sectional study design. The findings of this current study needs to be confirmed by prospective studies and randomized controlled trials if indicated. Third, in our study, MVD events were identified by reviewing medical records, not by universal complication screening. Although all inpatients of tertiary hospitals in China had detailed clinical examinations, some MVD cases may be missed. Fourth, lifestyle and social-economic factors, such as smoking, alcohol drinking habits, education levels, income and history of gout were not collected in the study, it was uncertain whether those factors could partially explain
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002
6
Y. Ren et al. / Journal of Diabetes and Its Complications xxx (xxxx) xxx–xxx
the observed associations reported in the study. Fifth, patients with abnormal kidney function, i.e., serum creatinine ≥125 μmol/l in male and ≥ 110 μmol/l in female or chronic kidney disease were not invited to participate in this survey and the endpoint of diabetes nephropathy only included mild diabetes nephropathy without declined renal function. So the observed associations in our study may underestimate the risk in the general patient population if those patients with declined renal function were included in the analysis. In summary, we found that co-presence of SUA ≥283 μmol/l and TBIL b11.5 μmol/l was associated with markedly increased risk of MVD in Chinese patients with T2DM. Isolated presence of either of them was not associated with increased risks of MVD. Co-presence of SUA ≥283 μmol/l and TBIL b11.5 μmol/l may be a potential predictor for MVD if our findings are confirmed in patients with T2DM, especially among low risk patients with T2DM. Acknowledgements This study was supported by a research grant from Novo Nordisk China. In addition, we thank fieldworkers involved in the survey for their efforts in data collection and quality control. References 1. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 2004;27:1047-53. 2. Fong DS, Aiello LP, Ferris III FL, Klein R. Diabetic retinopathy. Diabetes Care 2004;27: 2540-53. 3. Hwang SJ, Tsai JC, Chen HC. Epidemiology, impact and preventive care of chronic kidney disease in Taiwan. Nephrology (Carlton) 2010;15:3-9. 4. Chan KH, O'Connell RL, Sullivan DR, Hoffmann LS, Rajamani K, Whiting M, et al. Plasma total bilirubin levels predict amputation events in type 2 diabetes mellitus: the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study. Diabetologia 2013;56:724-36. 5. Liang CC, Lin PC, Lee MY, Chen SC, Shin SJ, Hsiao PJ, et al. Association of serum uric acid concentration with diabetic retinopathy and albuminuria in Taiwanese patients with type 2 diabetes mellitus. Int J Mol Sci 2016;17:1248. 6. Kang DH, Ha SK. Uric acid puzzle: dual role as anti-oxidantand pro-oxidant. Electrolyte Blood Press 2014;12:1-6. 7. Gillett MJ. International expert committee report on the role of the A1c assay in the diagnosis of diabetes. Clin Biochem Rev 2009;30:197-200. 8. Kuo CC, Weaver V, Fadrowski JJ, Lin YS, Guallar E, Navas-Acien A. Arsenic exposure, hyperuricemia, and gout in US adults. Environ Int 2015;76:32-40. 9. Takae K, Nagata M, Hata J, Mukai N, Hirakawa Y, Yoshida D, et al. Serum uric acid as a risk factor for chronic kidney disease in a Japanese community- the Hisayama study. Circ J 2016;80:1857-62. 10. Weiner DE, Tighiouart H, Elsayed EF, Griffith JL, Salem DN, Levey AS. Uric acid and incident kidney disease in the community. J Am Soc Nephrol 2008;19:1204-11. 11. Wu YQ, Li J, Xu YX, Wang YL, Luo YY, Hu DY, et al. Predictive value of serum uric acid on cardiovascular disease and all-cause mortality in urban Chinese patients. Chin Med J (Engl) 2010;123:1387-91. 12. Lee JJ, Yang IH, Kuo HK, Chung MS, Chen YJ, Chen CH, et al. Serum uric acid concentration is associated with worsening in severity of diabetic retinopathy among type 2 diabetic patients in Taiwan–a 3-year prospective study. Diabetes Res Clin Pract 2014;106:366-72. 13. Shani M, Vinker S, Dinour D, Leiba M, Twig G, Holtzman EJ, et al. High normal uric acid levels are associated with an increased risk of diabetes in lean, normoglycemic healthy women. J Clin Endocrinol Metab 2016;101:3772-8.
14. Baranano DE, Rao M, Ferris CD, Snyder SH. Biliverdin reductase: a major physiologic cytoprotectant. Proc Natl Acad Sci U S A 2002;99:16093-8. 15. Dekker D, Dorresteijn MJ, Pijnenburg M, Heemskerk S, Rasing-Hoogveld A, Burger DM, et al. The bilirubin-increasing drug atazanavir improves endothelial function in patients with type 2 diabetes mellitus. Arterioscler Thromb Vasc Biol 2011;31: 458-63. 16. Chin HJ, Song YR, Kim HS, Park M, Yoon HJ, Na KY, et al. The bilirubin level is negatively correlated with the incidence of hypertension in normotensive Korean population. J Korean Med Sci 2009;24:S50-6. [Suppl.]. 17. Hwang HJ, Kim SH. Inverse relationship between fasting direct bilirubin and metabolic syndrome in Korean adults. Clin Chim Acta 2010;411:1496-501. 18. Jung CH, Lee MJ, Kang YM, Hwang JY, Jang JE, Leem J, et al. Higher serum bilirubin level as a protective factor for the development of diabetes in healthy Korean men: a 4 year retrospective longitudinal study. Metabolism 2014;63:87-93. 19. Kang SJ, Kim D, Park HE, Chung GE, Choi SH, Choi SY, et al. Elevated serum bilirubin levels are inversely associated with coronary artery atherosclerosis. Atherosclerosis 2013;230:242-8. 20. Kim ES, Mo EY, Moon SD, Han JH. Inverse association between serum bilirubin levels and arterial stiffness in Korean women with type 2 diabetes. PLoS One 2014;9, e109251. 21. Sekioka R, Tanaka M, Nishimura T, Itoh H. Serum total bilirubin concentration is negatively associated with increasing severity of retinopathy in patients with type 2 diabetes mellitus. J Diabetes Complications 2015;29:218-21. 22. Troughton JA, Woodside JV, Young IS, Arveiler D, Amouyel P, Ferrieres J, et al. Bilirubin and coronary heart disease risk in the Prospective Epidemiological Study of Myocardial Infarction (PRIME). Eur J Cardiovasc Prev Rehabil 2007;14:79-84. 23. Ren Y, Jin N, Hong T, Mu Y, Guo L, Ji Q, et al. Interactive effect of serum uric acid and total bilirubin for cardiovascular disease in Chinese patients with type 2 diabetes. Sci Rep 2016;6:36437. 24. Chandy A, Pawar B, John M, Isaac R. Association between diabetic nephropathy and other diabetic microvascular and macrovascular complications. Saudi J Kidney Dis Transpl 2008;19:924-8. 25. Pradeepa R, Anjana RM, Unnikrishnan R, Ganesan A, Mohan V, Rema M. Risk factors for microvascular complications of diabetes among south Indian subjects with type 2 diabetes–the Chennai Urban Rural Epidemiology Study (CURES) Eye Study-5. Diabetes Technol Ther 2010;12:755-61. 26. Harrell F. Regression Modelling Strategies with Applications to Linear Models, Logistic Regression, and Survival Analysis. New York: Spinger-Varlag New York, Inc.. 2001. 27. Yang X, So WY, Ma RC, Kong AP, Lee HM, Yu LW, et al. Low HDL cholesterol, metformin use, and cancer risk in type 2 diabetes: the Hong Kong diabetes registry. Diabetes Care 2011;34:375-80. 28. Momeni A, Shahidi S, Seirafian S, Taheri S, Kheiri S. Effect of allopurinol in decreasing proteinuria in type 2 diabetic patients. Iran J Kidney Dis 2010;4:128-32. 29. De Cosmo S, Viazzi F, Pacilli A, Giorda C, Ceriello A, Gentile S, et al. Serum uric acid and risk of CKD in type 2 diabetes. Clin J Am Soc Nephrol 2015;10:1921-9. 30. Hamamoto S, Kaneto H, Kamei S, Shimoda M, Tawaramoto K, Kanda-Kimura Y, et al. Low bilirubin levels are an independent risk factor for diabetic retinopathy and nephropathy in Japanese patients with type 2 diabetes. Diabetes Metab 2015;41: 429-31. 31. Najam SS, Sun J, Zhang J, Xu M, Lu J, Sun K, et al. Serum total bilirubin levels and prevalence of diabetic retinopathy in a Chinese population. J Diabetes 2014;6:221-7. 32. Zhu B, Wu X, Ning K, Jiang F, Zhang L. The negative relationship between bilirubin level and diabetic retinopathy: a meta-analysis. PLoS One 2016;11, e0161649. 33. Toya K, Babazono T, Hanai K, Uchigata Y. Association of serum bilirubin levels with development and progression of albuminuria, and decline in estimated glomerular filtration rate in patients with type 2 diabetes mellitus. J Diabetes Investig 2014;5: 228-35. 34. Mene P, Punzo G. Uric acid: bystander or culprit in hypertension and progressive renal disease? J Hypertens 2008;26:2085-92. 35. Vitek L. The role of bilirubin in diabetes, metabolic syndrome, and cardiovascular diseases. Front Pharmacol 2012;3:55. 36. Rocuts F, Zhang X, Yan J, Yue Y, Thomas M, Bach FH, et al. Bilirubin promotes de novo generation of T regulatory cells. Cell Transplant 2010;19:443-51.
Please cite this article as: Ren Y, et al. Interactive effect of serum uric acid and total bilirubin for micro-vascular disease of type 2 diabetes in China. (2018), https://doi.org/10.1016/j.jdiacomp.2018.09.002