Clinica Chimica Acta 448 (2015) 232–237
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Insulin resistance is not independently associated with chronic kidney disease in Chinese population: A population-based cross-sectional study Ce Jing a,1, Shaoyong Xu a,1, Jie Ming a, Jing Cai a, Rong Zhang a, Han Shen a, Wenying Yang b, Qiuhe Ji a,⁎, on behalf of the China National Diabetes and Metabolic Disorders Study Group 2 a b
Department of Endocrinology, Xijing Hospital, Fourth Military Medical University, Xi'an, China Department of Endocrinology, China–Japan Friendship Hospital, Beijing, China
a r t i c l e
i n f o
Article history: Received 18 June 2015 Received in revised form 8 July 2015 Accepted 13 July 2015 Available online 14 July 2015 Keywords: Insulin resistance Metabolic syndrome Chronic kidney disease Chinese adults Epidemiology and outcomes
a b s t r a c t Background: Metabolic syndrome (MS) may modify the association of insulin resistance (IR) with chronic kidney disease (CKD), but the relevant studies were not enough. We evaluated whether IR is independently associated with CKD in Chinese population. Methods: The data were from 2007–2008 China National Diabetes and Metabolic Disorders Study. CKD was defined as an estimated glomerular filtration rate b 60 ml/min/1.73 m2. IR was evaluated by using the homeostatic model assessment (HOMA-IR). Results: A total of 11,143 individuals were included. Participants in the higher quartiles of HOMA-IR tended to have higher prevalence of CKD in general population (P b 0.001). However, there was no significant difference among the quartiles of HOMA-IR in population without MS (P = 0.288). In general population, the adjusted odds ratio of CKD was 1.183 (95% CI: 0.838–1.670), 1.543 (95% CI: 1.103–2.158), and 1.549 (95% CI: 1.079– 2.223) in the second, third and fourth quartile of HOMA-IR relative to the lowest quartile, while the odds ratios in population without MS showed no significance in all higher quartiles. Conclusions: IR was not an independently significant predictor of CKD in Chinese population, and MS may contribute greatly to the association between IR and CKD. © 2015 Elsevier B.V. All rights reserved.
1. Introduction China with a large population has a high prevalence of chronic kidney disease (CKD), which is estimated to be 10.8% and translated approximately a total of 119.5 million patients [1]. The increasing incidence of obesity, hypertension, and type 2 diabetes mellitus, coupled with aging population, will worsen the burden of CKD, which has substantial socioeconomic and public health consequence [2]. Insulin resistance (IR) is defined clinically in terms of the failure of insulin to maintain glucose homeostasis, which is indicated to be associated with CKD [3]. For example, end-stage renal disease is
Abbreviations: CKD, chronic kidney disease; MS, metabolic syndrome; IR, insulin resistance; SCr, serum creatinine; HDL-c, high-density lipoprotein cholesterol; CNY, China Yuan; NECP-ATP-III, National Cholesterol Education Program Adult Treatment Panel III; MDRD, Modification of Diet in Renal Disease. ⁎ Corresponding author at: Department of Endocrinology, Xijing Hospital, Fourth Military Medical University, 169 Changle Road West, Xi'an 710032, China. E-mail address: [email protected] (Q. Ji). 1 CJ and SX contributed equally to this work. 2 Members of the China National Diabetes and Metabolic Disorders Study Group are listed in Appendix S1.
http://dx.doi.org/10.1016/j.cca.2015.07.013 0009-8981/© 2015 Elsevier B.V. All rights reserved.
characterized by IR, and the severity of IR was is correlated with glomerular filtration rate (GFR) [4,5]. Meanwhile, IR is believed to play a central role in the pathogenesis of metabolic syndrome (MS), which is characterized as a cluster of metabolic abnormalities including obesity, hyperglycemia, hypertension and dyslipidemia [6,7]. IR and MS are not synonymous, despite there is significant overlap between the two terms. We previously reported that patients with MS had a 50% increase in the odds of CKD compared with individuals without MS [8]. Therefore, although several observational and prospective studies have indicated a relationship between IR and CKD in general or nondiabetic population [9–15], there were not enough studies concerning this relationship in populations without MS and these results were debatable [16,17]. It is of importance because MS may modify the association of IR with CKD, and the studies focusing the relationship between IR and CKD should better exclude patients with MS, rather than only exclude diabetic patients or adjust metabolic factors as variables. Given the above background, we utilized the data of the 2007–2008 China National Diabetes and Metabolic Disorders Study, aiming to explore the association between IR and CKD in Chinese population, and to evaluate whether IR is independently associated with CKD in population without MS.
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
233
2. Material and methods
2.3. Definitions
2.1. Study population
Metabolic syndrome was defined using the National Cholesterol Education Program Adult Treatment Panel III criteria (NECP-ATP-III) as three or more of the following five metabolic components [6]: 1) elevated waist circumference: ≥90 cm (males) or ≥ 80 cm (females); 2) increased triglycerides: ≥ 1.69 mmol/l or the use of lipid medications; 3) elevated blood pressure: systolic blood pressure ≥130 mm Hg, or diastolic blood pressure ≥85 mm Hg, or the use of antihypertensive medications; 4) elevated fasting glucose: ≥5.6 mmol/l or the use of diabetes medications; 5) reduced HDL-c: b1.04 mmol/l (male) or b1.29 mmol/l (female). The abbreviated equation developed by the Modification of Diet in Renal Disease (MDRD) study with modification for the Chinese population was used to calculate GFR [20]. Since most of our centers measured SCr on a Hitachi analyzer using the Jaffe's kinetic method, the following equation was adopted: 175 × (SCr × 0.01131) (mmol/l)−1.234 × age (year)−0.179 × (0.79 if female), which has been validated in the Chinese population and is also used by previous studies. CKD was defined as an estimated GFR b 60 ml/min/1.73 m2 according to the US National Kidney Foundation guidelines [21]. Diabetes was diagnosed based on Standard World Health Organization criteria [22]: fasting glucose ≥ 7.0 mmol/l or 2 h postprandial glucose ≥11.1 mmol/l, or self-reported use of diabetes medications. Hypertension was defined according to the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [23]: systolic blood pressure of 140 mm Hg or greater, diastolic blood pressure of 90 mm Hg or greater, or self-reported use of antihypertensive medications.
The data in present study were from the 2007–08 China National Diabetes and Metabolic Disorders Study, which was a nationwide population-based cross-sectional survey and conducted between June 2007 and May 2008. The details of the study have been previously reported [8,18]. In brief, a multi-stage stratified sampling method was used to select a nationwide representative sample of Chinese adults with age above 20 years. A total of 17 study group field centers participated in the study, and 54,240 individuals from the general population were selected and invited. Of those individuals, 87.3% (47,325 individuals: 18,976 men and 28,349 women) participated, and 85.2% (46,239 individuals: 18,419 men and 27,820 women) completed the study. Since kidney function and insulin measurement were not the primary expected outcomes, serum creatinine (SCr) and fasting serum insulin were not measured at all of the centers. Therefore, only 11,143 individuals (4419 men and 6724 women) were included as study population, which had complete data on SCr, fasting serum insulin, fasting plasma glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride level, serum high-density lipoprotein cholesterol (HDL-c), and waist circumference. The map of inclusive/exclusive centers can be seen in our previous report [8]. This study was approved by the institutional review boards or ethnics committees from all of the 17 participating centers. All the participants signed written informed consent prior to data collection. The 17 institutional review boards' approvals covered every participant in the study.
2.4. Statistical analysis 2.2. Data collection A standard questionnaire was designed and administered by trained physicians or nurses at local health stations or in community clinics to collect information on demographic characteristics, lifestyle risk factors, and personal medical history. Educational level was categorized as college level or above, secondary school, and elementary school or below. Yearly family income was categorized as b 10,000 China Yuan (CNY), 10,000–30,000 CNY and N30,000 CNY. Cigarette smoking was defined as a lifetime history of smoking at least 100 cigarettes. Frequency of drinking per week, favorite type of alcohol, and the amount of drinking per each occasion were recorded, and alcohol drinking was defined as the consumption of at least 30 g of alcohol per week for one year or more. Physical activity was defined as participating in moderate or vigorous activity for 30 min or more per day for at least 3 days a week. Body weight and height were measured without shoes and in light clothing, and body mass index calculated as weight in kilograms divided by the square of height in meters. Waist circumference was measured at the middle point between the costal margin and iliac crest. Blood pressure was measured using a standardized mercury sphygmomanometer in the sitting position after at least 5 min of rest; 2 consecutive readings of blood pressure were taken on the same arm and the mean of the 2 measures was used for analysis. Oral glucose tolerance test was performed on all subjects for the measurement of serum glucose and insulin. After at least 10 h of overnight fasting, participants with no history of diabetes were administered a standard 75 g glucose solution, while participants with a self-reported history of diabetes were given a steamed bun that contained approximately 80 g of complex carbohydrates for safety reasons. Fasting blood samples were also taken to measure serum triglyceride and HDL-c level. All laboratory measurements met a standardization and certification program [18]. IR was evaluated by using the homeostatic model assessment of insulin resistance (HOMA-IR) based on the following formula: fasting serum insulin (μU/ml) × fasting plasma glucose (mmol/l)/22.5 [19].
Data were analyzed using SPSS 18.0 (SPSS Inc.). Data were expressed as the mean ± SD, median with interquartile range, or percentage as suitable. Comparisons between groups were analyzed by t-test or Mann–Whitney U-test for measurement data, and χ2 test for enumeration data. We used univariate and multivariate logistics analysis to determine odds ratios (ORs) for presence of CKD according to the quartiles of HOMA-IR (it was not normally distributed and thus Log transformed). The lowest quartile of HOMA-IR was used as reference. Odds ratios and 95% confidence intervals (95% CI) were calculated using a forward stepwise method. The covariables in the multivariate analysis were age, gender, ethnics, educational level, yearly family income, cigarette smoking, alcohol drinking, physical activities, waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and high-density lipoprotein cholesterol. Data analyses were conducted firstly in general population to explore the association between IR and CKD, then in population without diabetes, population without hypertension, and population without MS to evaluate whether IR is independently associated with CKD. In addition, tertile analyses based on HOMA-IR were conducted as sensitivity analyses to test the robustness of the primary findings. A P b 0.05 was considered statistically significant. 3. Results The characteristics of study population according to the quartiles of HOMA-IR were described in Table 1. A total of 11,143 participants were included as study population. The mean age was 46.2 years; 39.7% of participants were men and 98.9% were Han ethnics. The mean HOMA-IR was 1.76 ± 1.14, with range of 0.38–9.99. The overall proportion of diabetes, hypertension and MS was 9.4%, 29.8%, and 29.3%, separately. Participants who had higher quartile of HOMA-IR were more likely to have higher level of metabolic parameters (e.g., blood pressure, fasting glucose level, serum triglyceride level and
234
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
Table 1 Characteristics of study population according to the quartiles of HOMA-IR. Variable
HOMA-IR Range Mean ± SD Age, year Gender (men), n (%) Ethnics (Han), n (%) Educational level, n (%) Elementary school or below Secondary school College or above Yearly family income, n (%) 10,000 CNY below 10,000–30,000 CNY 30,000 CNY above Cigarette smoking, n (%) Alcohol drinking, n (%) Physical activities, n (%) Waist circumference, cm Body mass index, kg/m2 Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Fasting glucose level, mmol/l 2 h postprandial glucose, mmol/l Serum triglyceride level, mmol/l Serum HDL-c level, mmol/l Serum creatinine level, μmol/l eGFR, ml/min per 1.73 m2 Hypertension, n (%) Diabetes, n (%) Metabolic syndrome, n (%)
Total
Q1
Q2
Q3
Q4
(n = 11,143)
(n = 2780)
(n = 2788)
(n = 2788)
(n = 2787)
0.38–9.99 1.76 ± 1.14 46.2 ± 13.6 4419 (39.7) 10,873 (98.9)
b1.05 0.83 ± 0.15 44.8 ± 13.8 1129 (40.6) 2690 (98.7)
1.05–1.46 1.25 ± 0.12 45.6 ± 13.5 1095 (39.3) 2722 (98.8)
1.46–2.09 1.74 ± 0.18 46.0 ± 13.4 1043 (37.4) 2725 (98.8)
N2.09 3.25 ± 1.34 48.2 ± 13.6 1152 (41.3) 2736 (99.5)
2235 (20.1) 6774 (61.3) 2040 (18.5)
566 (20.6) 1656 (60.2) 529 (19.2)
527 (19.1) 1735 (62.8) 502 (18.2)
540 (19.5) 1721 (62.2) 507 (18.3)
602 (21.8) 1662 (60.1) 502 (18.1)
3613 (34.4) 4244 (40.4) 2642 (25.2) 2603 (23.4) 2569 (23.1) 5475 (49.2) 81.8 ± 11.1 24.4 ± 3.8 121.7 ± 19.3 77.5 ± 11.2 5.3 ± 1.3 6.9 ± 3.3 1.6 ± 1.2 1.3 ± 0.3 71.0 ± 19.5 97.1 (80.8–115.7) 3320 (29.8) 1049 (9.4) 3262 (29.3)
944 (36.5) 1009 (39.0) 631 (24.4) 764 (27.5) 728 (26.2) 1257 (45.3) 77.9 ± 10.1 22.9 ± 3.2 117.5 ± 18.4 75.0 ± 10.7 4.8 ± 0.6 5.9 ± 1.7 1.3 ± 0.9 1.4 ± 0.3 70.5 ± 17.9 98.1(81.8–116.7) 575 (20.7) 52 (1.9) 322 (11.6)
949 (35.9) 1074 (40.6) 623 (23.5) 649 (23.3) 629 (22.6) 1393 (50.0) 80.0 ± 10.4 23.7 ± 3.5 119.7 ± 18.7 76.3 ± 10.8 5.0 ± 0.7 6.2 ± 2.1 1.4 ± 0.9 1.4 ± 0.3 71.1 ± 18.9 96.7 (80.7–114.5) 688 (24.7) 116 (4.2) 536 (19.2)
886 (33.8) 1068 (40.7) 669 (25.5) 580 (20.8) 583 (20.9) 1427 (51.2) 82.3 ± 10.5 24.6 ± 3.6 122.3 ± 19.2 78.0 ± 11.0 5.3 ± 0.9 6.7 ± 2.7 1.6 ± 1.2 1.3 ± 0.3 71.2 ± 18.9 96.4 (80.2–114.6) 861 (30.9) 222 (8.0) 861 (30.9)
834 (31.5) 1093 (41.3) 719 (27.2) 610 (21.9) 629 (22.6) 1398 (50.3) 87.2 ± 11.1 26.3 ± 3.9 127.2 ± 19.6 80.8 ± 11.4 6.1 ± 1.1 8.7 ± 4.8 2.1 ± 1.5 1.3 ± 0.3 71.2 ± 21.4 96.9 (80.1–116.5) 1196 (42.9) 659 (23.6) 1543 (55.4)
P
b0.001 0.016 0.018 NS
0.002
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 NS NS b0.001 b0.001 b0.001
HOMA-IR: homeostasis model assessment of insulin resistance; CNY: China Yuan.
serum HDL-c level), and higher proportion of diabetes, hypertension and MS. The overall prevalence of CKD in general population was 3.7%. Participants in the higher quartiles of HOMA-IR tended to have higher prevalence of CKD in general population (P b 0.001) (Fig. 1A), participants without diabetes (P = 0.002) (Fig. 1B) and participants without hypertension (P b 0.001) (Fig. 1C). However, there was no significant difference among the four quartiles of HOMA-IR in participants without MS (P = 0.288) (Fig. 1D). Logistics analyses showed that in general participants the adjusted odds ratio of CKD was 1.183 (95% CI: 0.838–1.670), 1.543 (95% CI: 1.103–2.158), and 1.549 (95% CI: 1.079–2.223) in the second, third and fourth quartile of HOMA-IR relative to the lowest quartile after adjustment for age, gender, ethnics, educational level, yearly family income, cigarette smoking, alcohol drinking, physical activities, waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and high-density lipoprotein cholesterol. A positive trend was observed between increasing HOMA-IR and prevalence of CKD (P for trend b 0.001) (Table 2). Table 3 showed the crude and multivariate-adjusted ORs of CKD associated with HOMA-IR in participants without diabetes and participants without hypertension. These results were consistent with analyses in general population. However, in participants without MS the multivariate-adjusted ORs of CKD associated with HOMA-IR showed no significance in the second (OR: 1.183, 95% CI: 0.838– 1.670), third (OR: 1.543, 95% CI: 1.103–2.158) and fourth (OR: 1.549, 95% CI: 1.079–2.223) quartile of HOMA-IR relative to the lowest quartile. In addition, sensitivity analyses showed that the primary findings were not influenced in tertile analyses and the results were consistent (data not shown).
4. Discussion In this cross-sectional study of a large nationwide cohort of Chinese population, we provided the evidence that IR was significantly associated with the presence of CKD, defined as an estimated GFR b60 ml/min/1.73 m2, in general population. However, the association lost its significance after excluding patients with MS, indicating that MS may contribute greatly to the association between IR and CKD, and IR was not an independently significant predictor of CKD in Chinese population. Similar with ours, most previous studies have suggested an association between IR and increased risk of CKD. For example, Chen et al. analyzed the data of the Third National Health and Nutrition Examination Survey which included 6453 US adults without diabetes. The results showed that the prevalence of CKD, defined as an estimated GFR b60 ml/min/1.73 m2, was significantly higher with increasing levels of HOMA-IR, and the odds ratio of CKD for the highest compared with the lowest quartile was 2.65 (95% CI: 1.25–5.62, P = 0.008) [15]. In the Health, Aging and Body Composition study which included older individuals without diabetes, eGFR (odds ratio per 10 ml/min/1.73 m2: 0.92, 95% CI: 0.87–0.98) and CKD (1.41, 95% CI: 1.04–1.92) were both independently associated with HOMA-IR [13]. Similarly, Mykkanen et al. described a strict and independent association between IR, estimated by a frequently sampled intravenous glucose tolerance test, and CKD, defined as having microalbuminuria, in non-diabetic individuals [24]. Diabetes has been identified as a major underlying cause of kidney dysfunction [25]. Several cohort studies have indicated an association between diabetes and the increased risk of end-stage renal disease [26,27]. Therefore, most studies focusing the relationship between IR and CKD excluded individuals with diabetes. However, any insulin resistance-related abnormalities may contribute to kidney dysfunction
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
235
Fig. 1. Prevalence of chronic kidney disease in general population (A), population without diabetes (B), population without hypertension (C), and population without metabolic syndrome (D) according to the quartiles of HOMA-IR. HOMA-IR: homeostasis model assessment of insulin resistance; chronic kidney Disease (CKD) was defined as an estimated GFR b60 ml/min/1.73 m2.
[28]. Each single component of constellation of abnormalities clustering under the definition of metabolic syndrome may in fact play a role in kidney damage [28]. For example, abnormal blood pressure may lead to CKD mainly due to intraglomerular hypertension [29]. Generally speaking, when evaluating the effect of one factor to an association, the best way is subgroup analysis, rather than multivariable analysis. Therefore, in order to investigate the influence of MS in the association between IR and CKD, a subgroup analysis by excluding population with MS was conducted. It is known that the group of patients without MS could have 0–2 characteristics of MS and all five characteristics contribute with a different weight to CKD prevalence. So in the subgroup analyses, all the related metabolic variables (e.g. waist circumference, blood glucose, blood pressure) were still adjusted for. However, in the subgroup analysis which excluded participants with MS, we found that the association between IR and CKD lost its significance. Our findings were different from those in Jang's study [17]. Jang et al. conducted a cohort study which included 60,047 participants to evaluate the effect of IR on the 2-year development of CKD, defined as having albuminuria. The results showed that 5th quintile of HOMA-IR compared to 1st quintile of HOMA-IR was associated with the
development of albuminuria in subjects without MS (OR: 1.51, 95% CI: 1.09–2.11, P = 0.014). However, the associations were not significant in other quintiles of HOMA-IR compared to the 1st quintile [17]. Similarly, our results were in accord with Park's study [16], which included 17,157 subjects above 20 years old who underwent voluntary health check-ups. The results showed that while HOMA-IR index values were higher in lower eGFR groups compared with higher eGFR groups (P b 0.001), no meaningful differences in HOMA-IR values according to eGFR groups were observed in subjects without MS [16]. Therefore, our study showed that IR is not independently associated with CKD in Chinese population; our study also indicated that MS, as a constellation of metabolic abnormalities, may contribute greatly to and modify the association of IR with CKD, despite individual component may not influence the significance of the association. The major strength of our study was the nationwide populationbased representatives, so the findings could be generalized to Chinese population. However, several limitations should be addressed. Firstly, the cross-sectional study design limited to draw conclusion regarding the causality among IR, MS and CKD. Our results could not answer the questions whether IR and MS contributed to the initiation or
Table 2 Crude and multivariate-adjusted odds ratio of chronic kidney disease associated with HOMA-IR in general population. Variable
n/N % ± SE Model 0 Model 1 Model 2 Model 3
Q1
Q2
OR
OR (95% CI)
72/2780 2.6 ± 0.3 1.000 (ref.) 1.000 (ref.) 1.000 (ref.) 1.000 (ref.)
93/2788 3.3 ± 0.3 1.288 (0.922–1.800) 1.209 (0.861–1.697) 1.165 (0.829–1.639) 1.183 (0.838–1.670)
Q3 P
OR (95% CI)
NS NS NS NS
117/2788 4.2 ± 0.4 1.724 (1.255–2.367) 1.581 (1.146–2.182) 1.519 (1.097–2.102) 1.543 (1.103–2.158)
Q4 P
OR (95% CI)
0.001 0.005 0.012 0.011
133/2787 4.8 ± 0.4 1.933 (1.416–2.639) 1.573 (1.146–2.159) 1.563 (1.135–2.151) 1.549 (1.079–2.223)
P for trend P b0.001 b0.001 0.005 0.006 0.018
HOMA-IR: homeostasis model assessment of insulin resistance. Model 0: unadjusted; Model 1: adjusted for age, gender and ethnics; Model 2: Model 1 plus further adjusted for educational level, yearly family income, cigarette smoking, alcohol drinking, and physical activities; Model 3: Model 2 plus further adjusted for waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and HDL cholesterol.
236
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
Table 3 Crude and multivariate-adjusted odds ratio of chronic kidney disease associated with HOMA-IR in population without metabolic syndrome. Variable
Q1
Q2
OR
OR (95% CI)
Participants without diabetes n/N 70/2728 % ± SE 2.6 ± 0.3 Crude OR 1.000 (ref.) Adjusted OR 1.000 (ref.)
89/2672 3.3 ± 0.3 1.286 (0.913–1.811) 1.226 (0.863–1.734)
Participants without hypertension n/N 44/2205 % ± SE 2.0 ± 0.3 Crude OR 1.000 (ref.) Adjusted OR 1.000 (ref.)
56/2100 2.7 ± 0.3 1.291 (0.848–1.966) 1.212 (0.788–1.864)
Participants without metabolic syndrome n/N 56/2458 65/2252 % ± SE 2.3 ± 0.3 2.9 ± 0.3 Crude OR 1.000 (ref.) 1.214 (0.829–1.776) Adjusted OR 1.000 (ref.) 1.178 (0.794–1.747)
Q3
Q4
P
OR (95% CI)
NS NS
105/1927 4.1 ± 0.3 1.708 (1.232–2.369) 1.622 (1.151–2.286)
NS NS
60/1927 3.1 ± 0.3 1.587 (1.052–2.393) 1.485 (0.972–2.271)
NS NS
59/1927 3.1 ± 0.3 1.382 (0.941–2.030) 1.366 (0.909–2.052)
P
OR (95% CI)
0.001 0.006
85/2128 4.0 ± 0.4 1.579 (1.119–2.229) 1.459 (1.005–2.120)
0.028 NS
66/1591 4.1 ± 0.4 2.000 (1.329–3.009) 1.717 (1.112–2.651)
NS NS
40/1244 3.2 ± 0.4 1.256 (0.806–1.957) 1.314 (0.803–2.151)
P for trend P 0.002 0.009 0.047 b0.001 0.001 0.015
NS NS NS
HOMA-IR: homeostasis model assessment of insulin resistance. Adjusted OR: adjusted for age, gender, ethnics, educational level, yearly family income, cigarette smoking, alcohol drinking, physical activities, waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and HDL cholesterol.
progression of CKD, or vice versa. Secondly, GFR was not directly measured, since the methodologies are complex and not feasible in a large epidemiologic study. We thus adopted a modified MDRD equation to assess GFR which showed lower bias and the higher accuracy in each stage of CKD in Chinese population [20]. Similarly, although the euglycemic clamping method is the gold standard for measuring IR, the technique is not feasible in a large epidemiologic study and HOMA-IR was thus used. Thirdly, although we had a relative large study sample, we noticed that the sample size of patients with CKD in the different quartiles were low. The study may not be powered to draw strong conclusions due to low power to demonstrate a difference. Fourthly, we did not have the data of albuminuria and the frequency of CKD was underestimated. The power of the findings would be more adequate if having the data of albuminuria. Finally, a large amount of participants were excluded in the present study, which would be expected to bias the study results to some extent. However, most of the participants were excluded based on centers; for individual included center, the study was well-designed and the response rate was estimated above 80%. In summary, by using a nationwide representative sample, we indicated that IR was not an independently significant predictor of CKD in Chinese population, and MS may contribute greatly to the association between IR and CKD. More studies, particular cohort studies, are required to confirm the present findings. Acknowledgment We thank all physicians and participants of the study, for their cooperation and generous participation. This study was supported by the Chinese Medical Association Foundation and the Chinese Diabetes Society. QJ was supported by the Xijing Hospital Foundation (Grant No. XJZT13Z04). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cca.2015.07.013. References [1] L. Zhang, F. Wang, L. Wang, et al., Prevalence of chronic kidney disease in China: a cross-sectional survey, Lancet 379 (2012) 815–822. [2] R.A. Nugent, S.F. Fathima, A.B. Feigl, D. Chyung, The burden of chronic kidney disease on developing nations: a 21st century challenge in global health, Nephron Clin. Pract. 118 (2011) c269–277.
[3] M.T. Liao, C.C. Sung, K.C. Hung, et al., Insulin resistance in patients with chronic kidney disease, J. Biomed. Biotechnol. 2012 (2012) 691369. [4] G. Bodlaj, J. Berg, R. Pichler, G. Biesenbach, Prevalence, severity and predictors of HOMA-estimated insulin resistance in diabetic and nondiabetic patients with endstage renal disease, J. Nephrol. 19 (2006) 607–612. [5] B. Satirapoj, O. Supasyndh, P. Phantana-Angkul, et al., Insulin resistance in dialysis versus non dialysis end stage renal disease patients without diabetes, J. Med. Assoc. Thai. 94 (Suppl. 4) (2011) S87–93. [6] K.G. Alberti, R.H. Eckel, S.M. Grundy, et al., Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity, Circulation 120 (2009) 1640–1645. [7] S.M. Grundy, J.I. Cleeman, S.R. Daniels, et al., Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement, Circulation 112 (2005) 2735–2752. [8] J. Ming, S. Xu, C. Yang, et al., Metabolic syndrome and chronic kidney disease in general Chinese adults: results from the 2007–08 China National Diabetes and Metabolic Disorders Study, Clin. Chim. Acta 430 (2014) 115–120. [9] B.R. Johns, A.C. Pao, S.H. Kim, Metabolic syndrome, insulin resistance and kidney function in non-diabetic individuals, Nephrol. Dial. Transplant. 27 (2012) 1410–1415. [10] T. Basturk, A. Unsal, Is insulin resistance a risk factor for the progression of chronic kidney disease? Kidney Blood Press. Res. 34 (2011) 111–115. [11] B. Satirapoj, O. Supasyndh, A. Boonyavarakul, L. Luesutthiviboon, P. Choovichian, The correlation of insulin resistance and renal function in non diabetic chronic kidney disease patients, J. Med. Assoc. Thail. 88 (Suppl. 3) (2005) S97–104. [12] H. Pham, C. Robinson-Cohen, M.L. Biggs, et al., Chronic kidney disease, insulin resistance, and incident diabetes in older adults, Clin. J. Am. Soc. Nephrol. 7 (2012) 588–594. [13] M. Landau, M. Kurella-Tamura, M.G. Shlipak, et al., Correlates of insulin resistance in older individuals with and without kidney disease, Nephrol. Dial. Transplant. 26 (2011) 2814–2819. [14] S. Kobayashi, K. Maesato, H. Moriya, T. Ohtake, T. Ikeda, Insulin resistance in patients with chronic kidney disease, Am. J. Kidney Dis. 45 (2005) 275–280. [15] J. Chen, P. Muntner, L.L. Hamm, et al., Insulin resistance and risk of chronic kidney disease in nondiabetic US adults, J. Am. Soc. Nephrol. 14 (2003) 469–477. [16] J.H. Park, S.W. Oh, S.Y. Ahn, et al., Decreased estimated glomerular filtration rate is not directly related to increased insulin resistance, Diabetes Res. Clin. Pract. 99 (2013) 366–371. [17] C.M. Jang, Y.Y. Hyun, K.B. Lee, H. Kim, Insulin resistance is associated with the development of albuminuria in Korean subjects without diabetes, Endocrine 48 (2015) 203–210. [18] W. Yang, J. Lu, J. Weng, et al., Prevalence of diabetes among men and women in China, N. Engl. J. Med. 362 (2010) 1090–1101. [19] D.R. Matthews, J.P. Hosker, A.S. Rudenski, et al., Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man, Diabetologia 28 (1985) 412–419. [20] Y.C. Ma, L. Zuo, J.H. Chen, et al., Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease, J. Am. Soc. Nephrol. 17 (2006) 2937–2944. [21] K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification, Am. J. Kidney Dis. 39 (2002) S1–266. [22] K.G. Alberti, P.Z. Zimmet, Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237 mellitus provisional report of a WHO consultation, Diabet. Med. 15 (1998) 539–553. [23] A.V. Chobanian, G.L. Bakris, H.R. Black, et al., Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure, Hypertension 42 (2003) 1206–1252. [24] L. Mykkanen, D.J. Zaccaro, L.E. Wagenknecht, et al., Microalbuminuria is associated with insulin resistance in nondiabetic subjects: the insulin resistance atherosclerosis study, Diabetes 47 (1998) 793–800. [25] S.B. Ghaderian, F. Hayati, S. Shayanpour, S.S. Beladi Mousavi, Diabetes and end-stage renal disease; a review article on new concepts, J. Renal Inj. Prev. 4 (2015) 28–33.
237
[26] W.F. Keane, B.M. Brenner, D. de Zeeuw, et al., The risk of developing end-stage renal disease in patients with type 2 diabetes and nephropathy: the RENAAL study, Kidney Int. 63 (2003) 1499–1507. [27] F.L. Brancati, P.K. Whelton, B.L. Randall, et al., Risk of end-stage renal disease in diabetes mellitus: a prospective cohort study of men screened for MRFIT. Multiple Risk Factor Intervention Trial, JAMA 278 (1997) 2069–2074. [28] Y.J. Sheen, W.H. Sheu, Metabolic syndrome and renal injury, Cardiol. Res. Pract. 2011 (2011) 567389. [29] M.E. Hall, J.M. do Carmo, A.A. da Silva, et al., Obesity, hypertension, and chronic kidney disease, Int. J. Nephrol. Renov. Dis. 7 (2014) 75–88.
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Insulin resistance is not independently associated with chronic kidney disease in Chinese population: A population-based cross-sectional study Ce Jing a,1, Shaoyong Xu a,1, Jie Ming a, Jing Cai a, Rong Zhang a, Han Shen a, Wenying Yang b, Qiuhe Ji a,⁎, on behalf of the China National Diabetes and Metabolic Disorders Study Group 2 a b
Department of Endocrinology, Xijing Hospital, Fourth Military Medical University, Xi'an, China Department of Endocrinology, China–Japan Friendship Hospital, Beijing, China
a r t i c l e
i n f o
Article history: Received 18 June 2015 Received in revised form 8 July 2015 Accepted 13 July 2015 Available online 14 July 2015 Keywords: Insulin resistance Metabolic syndrome Chronic kidney disease Chinese adults Epidemiology and outcomes
a b s t r a c t Background: Metabolic syndrome (MS) may modify the association of insulin resistance (IR) with chronic kidney disease (CKD), but the relevant studies were not enough. We evaluated whether IR is independently associated with CKD in Chinese population. Methods: The data were from 2007–2008 China National Diabetes and Metabolic Disorders Study. CKD was defined as an estimated glomerular filtration rate b 60 ml/min/1.73 m2. IR was evaluated by using the homeostatic model assessment (HOMA-IR). Results: A total of 11,143 individuals were included. Participants in the higher quartiles of HOMA-IR tended to have higher prevalence of CKD in general population (P b 0.001). However, there was no significant difference among the quartiles of HOMA-IR in population without MS (P = 0.288). In general population, the adjusted odds ratio of CKD was 1.183 (95% CI: 0.838–1.670), 1.543 (95% CI: 1.103–2.158), and 1.549 (95% CI: 1.079– 2.223) in the second, third and fourth quartile of HOMA-IR relative to the lowest quartile, while the odds ratios in population without MS showed no significance in all higher quartiles. Conclusions: IR was not an independently significant predictor of CKD in Chinese population, and MS may contribute greatly to the association between IR and CKD. © 2015 Elsevier B.V. All rights reserved.
1. Introduction China with a large population has a high prevalence of chronic kidney disease (CKD), which is estimated to be 10.8% and translated approximately a total of 119.5 million patients [1]. The increasing incidence of obesity, hypertension, and type 2 diabetes mellitus, coupled with aging population, will worsen the burden of CKD, which has substantial socioeconomic and public health consequence [2]. Insulin resistance (IR) is defined clinically in terms of the failure of insulin to maintain glucose homeostasis, which is indicated to be associated with CKD [3]. For example, end-stage renal disease is
Abbreviations: CKD, chronic kidney disease; MS, metabolic syndrome; IR, insulin resistance; SCr, serum creatinine; HDL-c, high-density lipoprotein cholesterol; CNY, China Yuan; NECP-ATP-III, National Cholesterol Education Program Adult Treatment Panel III; MDRD, Modification of Diet in Renal Disease. ⁎ Corresponding author at: Department of Endocrinology, Xijing Hospital, Fourth Military Medical University, 169 Changle Road West, Xi'an 710032, China. E-mail address: [email protected] (Q. Ji). 1 CJ and SX contributed equally to this work. 2 Members of the China National Diabetes and Metabolic Disorders Study Group are listed in Appendix S1.
http://dx.doi.org/10.1016/j.cca.2015.07.013 0009-8981/© 2015 Elsevier B.V. All rights reserved.
characterized by IR, and the severity of IR was is correlated with glomerular filtration rate (GFR) [4,5]. Meanwhile, IR is believed to play a central role in the pathogenesis of metabolic syndrome (MS), which is characterized as a cluster of metabolic abnormalities including obesity, hyperglycemia, hypertension and dyslipidemia [6,7]. IR and MS are not synonymous, despite there is significant overlap between the two terms. We previously reported that patients with MS had a 50% increase in the odds of CKD compared with individuals without MS [8]. Therefore, although several observational and prospective studies have indicated a relationship between IR and CKD in general or nondiabetic population [9–15], there were not enough studies concerning this relationship in populations without MS and these results were debatable [16,17]. It is of importance because MS may modify the association of IR with CKD, and the studies focusing the relationship between IR and CKD should better exclude patients with MS, rather than only exclude diabetic patients or adjust metabolic factors as variables. Given the above background, we utilized the data of the 2007–2008 China National Diabetes and Metabolic Disorders Study, aiming to explore the association between IR and CKD in Chinese population, and to evaluate whether IR is independently associated with CKD in population without MS.
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
233
2. Material and methods
2.3. Definitions
2.1. Study population
Metabolic syndrome was defined using the National Cholesterol Education Program Adult Treatment Panel III criteria (NECP-ATP-III) as three or more of the following five metabolic components [6]: 1) elevated waist circumference: ≥90 cm (males) or ≥ 80 cm (females); 2) increased triglycerides: ≥ 1.69 mmol/l or the use of lipid medications; 3) elevated blood pressure: systolic blood pressure ≥130 mm Hg, or diastolic blood pressure ≥85 mm Hg, or the use of antihypertensive medications; 4) elevated fasting glucose: ≥5.6 mmol/l or the use of diabetes medications; 5) reduced HDL-c: b1.04 mmol/l (male) or b1.29 mmol/l (female). The abbreviated equation developed by the Modification of Diet in Renal Disease (MDRD) study with modification for the Chinese population was used to calculate GFR [20]. Since most of our centers measured SCr on a Hitachi analyzer using the Jaffe's kinetic method, the following equation was adopted: 175 × (SCr × 0.01131) (mmol/l)−1.234 × age (year)−0.179 × (0.79 if female), which has been validated in the Chinese population and is also used by previous studies. CKD was defined as an estimated GFR b 60 ml/min/1.73 m2 according to the US National Kidney Foundation guidelines [21]. Diabetes was diagnosed based on Standard World Health Organization criteria [22]: fasting glucose ≥ 7.0 mmol/l or 2 h postprandial glucose ≥11.1 mmol/l, or self-reported use of diabetes medications. Hypertension was defined according to the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure [23]: systolic blood pressure of 140 mm Hg or greater, diastolic blood pressure of 90 mm Hg or greater, or self-reported use of antihypertensive medications.
The data in present study were from the 2007–08 China National Diabetes and Metabolic Disorders Study, which was a nationwide population-based cross-sectional survey and conducted between June 2007 and May 2008. The details of the study have been previously reported [8,18]. In brief, a multi-stage stratified sampling method was used to select a nationwide representative sample of Chinese adults with age above 20 years. A total of 17 study group field centers participated in the study, and 54,240 individuals from the general population were selected and invited. Of those individuals, 87.3% (47,325 individuals: 18,976 men and 28,349 women) participated, and 85.2% (46,239 individuals: 18,419 men and 27,820 women) completed the study. Since kidney function and insulin measurement were not the primary expected outcomes, serum creatinine (SCr) and fasting serum insulin were not measured at all of the centers. Therefore, only 11,143 individuals (4419 men and 6724 women) were included as study population, which had complete data on SCr, fasting serum insulin, fasting plasma glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride level, serum high-density lipoprotein cholesterol (HDL-c), and waist circumference. The map of inclusive/exclusive centers can be seen in our previous report [8]. This study was approved by the institutional review boards or ethnics committees from all of the 17 participating centers. All the participants signed written informed consent prior to data collection. The 17 institutional review boards' approvals covered every participant in the study.
2.4. Statistical analysis 2.2. Data collection A standard questionnaire was designed and administered by trained physicians or nurses at local health stations or in community clinics to collect information on demographic characteristics, lifestyle risk factors, and personal medical history. Educational level was categorized as college level or above, secondary school, and elementary school or below. Yearly family income was categorized as b 10,000 China Yuan (CNY), 10,000–30,000 CNY and N30,000 CNY. Cigarette smoking was defined as a lifetime history of smoking at least 100 cigarettes. Frequency of drinking per week, favorite type of alcohol, and the amount of drinking per each occasion were recorded, and alcohol drinking was defined as the consumption of at least 30 g of alcohol per week for one year or more. Physical activity was defined as participating in moderate or vigorous activity for 30 min or more per day for at least 3 days a week. Body weight and height were measured without shoes and in light clothing, and body mass index calculated as weight in kilograms divided by the square of height in meters. Waist circumference was measured at the middle point between the costal margin and iliac crest. Blood pressure was measured using a standardized mercury sphygmomanometer in the sitting position after at least 5 min of rest; 2 consecutive readings of blood pressure were taken on the same arm and the mean of the 2 measures was used for analysis. Oral glucose tolerance test was performed on all subjects for the measurement of serum glucose and insulin. After at least 10 h of overnight fasting, participants with no history of diabetes were administered a standard 75 g glucose solution, while participants with a self-reported history of diabetes were given a steamed bun that contained approximately 80 g of complex carbohydrates for safety reasons. Fasting blood samples were also taken to measure serum triglyceride and HDL-c level. All laboratory measurements met a standardization and certification program [18]. IR was evaluated by using the homeostatic model assessment of insulin resistance (HOMA-IR) based on the following formula: fasting serum insulin (μU/ml) × fasting plasma glucose (mmol/l)/22.5 [19].
Data were analyzed using SPSS 18.0 (SPSS Inc.). Data were expressed as the mean ± SD, median with interquartile range, or percentage as suitable. Comparisons between groups were analyzed by t-test or Mann–Whitney U-test for measurement data, and χ2 test for enumeration data. We used univariate and multivariate logistics analysis to determine odds ratios (ORs) for presence of CKD according to the quartiles of HOMA-IR (it was not normally distributed and thus Log transformed). The lowest quartile of HOMA-IR was used as reference. Odds ratios and 95% confidence intervals (95% CI) were calculated using a forward stepwise method. The covariables in the multivariate analysis were age, gender, ethnics, educational level, yearly family income, cigarette smoking, alcohol drinking, physical activities, waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and high-density lipoprotein cholesterol. Data analyses were conducted firstly in general population to explore the association between IR and CKD, then in population without diabetes, population without hypertension, and population without MS to evaluate whether IR is independently associated with CKD. In addition, tertile analyses based on HOMA-IR were conducted as sensitivity analyses to test the robustness of the primary findings. A P b 0.05 was considered statistically significant. 3. Results The characteristics of study population according to the quartiles of HOMA-IR were described in Table 1. A total of 11,143 participants were included as study population. The mean age was 46.2 years; 39.7% of participants were men and 98.9% were Han ethnics. The mean HOMA-IR was 1.76 ± 1.14, with range of 0.38–9.99. The overall proportion of diabetes, hypertension and MS was 9.4%, 29.8%, and 29.3%, separately. Participants who had higher quartile of HOMA-IR were more likely to have higher level of metabolic parameters (e.g., blood pressure, fasting glucose level, serum triglyceride level and
234
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
Table 1 Characteristics of study population according to the quartiles of HOMA-IR. Variable
HOMA-IR Range Mean ± SD Age, year Gender (men), n (%) Ethnics (Han), n (%) Educational level, n (%) Elementary school or below Secondary school College or above Yearly family income, n (%) 10,000 CNY below 10,000–30,000 CNY 30,000 CNY above Cigarette smoking, n (%) Alcohol drinking, n (%) Physical activities, n (%) Waist circumference, cm Body mass index, kg/m2 Systolic blood pressure, mm Hg Diastolic blood pressure, mm Hg Fasting glucose level, mmol/l 2 h postprandial glucose, mmol/l Serum triglyceride level, mmol/l Serum HDL-c level, mmol/l Serum creatinine level, μmol/l eGFR, ml/min per 1.73 m2 Hypertension, n (%) Diabetes, n (%) Metabolic syndrome, n (%)
Total
Q1
Q2
Q3
Q4
(n = 11,143)
(n = 2780)
(n = 2788)
(n = 2788)
(n = 2787)
0.38–9.99 1.76 ± 1.14 46.2 ± 13.6 4419 (39.7) 10,873 (98.9)
b1.05 0.83 ± 0.15 44.8 ± 13.8 1129 (40.6) 2690 (98.7)
1.05–1.46 1.25 ± 0.12 45.6 ± 13.5 1095 (39.3) 2722 (98.8)
1.46–2.09 1.74 ± 0.18 46.0 ± 13.4 1043 (37.4) 2725 (98.8)
N2.09 3.25 ± 1.34 48.2 ± 13.6 1152 (41.3) 2736 (99.5)
2235 (20.1) 6774 (61.3) 2040 (18.5)
566 (20.6) 1656 (60.2) 529 (19.2)
527 (19.1) 1735 (62.8) 502 (18.2)
540 (19.5) 1721 (62.2) 507 (18.3)
602 (21.8) 1662 (60.1) 502 (18.1)
3613 (34.4) 4244 (40.4) 2642 (25.2) 2603 (23.4) 2569 (23.1) 5475 (49.2) 81.8 ± 11.1 24.4 ± 3.8 121.7 ± 19.3 77.5 ± 11.2 5.3 ± 1.3 6.9 ± 3.3 1.6 ± 1.2 1.3 ± 0.3 71.0 ± 19.5 97.1 (80.8–115.7) 3320 (29.8) 1049 (9.4) 3262 (29.3)
944 (36.5) 1009 (39.0) 631 (24.4) 764 (27.5) 728 (26.2) 1257 (45.3) 77.9 ± 10.1 22.9 ± 3.2 117.5 ± 18.4 75.0 ± 10.7 4.8 ± 0.6 5.9 ± 1.7 1.3 ± 0.9 1.4 ± 0.3 70.5 ± 17.9 98.1(81.8–116.7) 575 (20.7) 52 (1.9) 322 (11.6)
949 (35.9) 1074 (40.6) 623 (23.5) 649 (23.3) 629 (22.6) 1393 (50.0) 80.0 ± 10.4 23.7 ± 3.5 119.7 ± 18.7 76.3 ± 10.8 5.0 ± 0.7 6.2 ± 2.1 1.4 ± 0.9 1.4 ± 0.3 71.1 ± 18.9 96.7 (80.7–114.5) 688 (24.7) 116 (4.2) 536 (19.2)
886 (33.8) 1068 (40.7) 669 (25.5) 580 (20.8) 583 (20.9) 1427 (51.2) 82.3 ± 10.5 24.6 ± 3.6 122.3 ± 19.2 78.0 ± 11.0 5.3 ± 0.9 6.7 ± 2.7 1.6 ± 1.2 1.3 ± 0.3 71.2 ± 18.9 96.4 (80.2–114.6) 861 (30.9) 222 (8.0) 861 (30.9)
834 (31.5) 1093 (41.3) 719 (27.2) 610 (21.9) 629 (22.6) 1398 (50.3) 87.2 ± 11.1 26.3 ± 3.9 127.2 ± 19.6 80.8 ± 11.4 6.1 ± 1.1 8.7 ± 4.8 2.1 ± 1.5 1.3 ± 0.3 71.2 ± 21.4 96.9 (80.1–116.5) 1196 (42.9) 659 (23.6) 1543 (55.4)
P
b0.001 0.016 0.018 NS
0.002
b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 b0.001 NS NS b0.001 b0.001 b0.001
HOMA-IR: homeostasis model assessment of insulin resistance; CNY: China Yuan.
serum HDL-c level), and higher proportion of diabetes, hypertension and MS. The overall prevalence of CKD in general population was 3.7%. Participants in the higher quartiles of HOMA-IR tended to have higher prevalence of CKD in general population (P b 0.001) (Fig. 1A), participants without diabetes (P = 0.002) (Fig. 1B) and participants without hypertension (P b 0.001) (Fig. 1C). However, there was no significant difference among the four quartiles of HOMA-IR in participants without MS (P = 0.288) (Fig. 1D). Logistics analyses showed that in general participants the adjusted odds ratio of CKD was 1.183 (95% CI: 0.838–1.670), 1.543 (95% CI: 1.103–2.158), and 1.549 (95% CI: 1.079–2.223) in the second, third and fourth quartile of HOMA-IR relative to the lowest quartile after adjustment for age, gender, ethnics, educational level, yearly family income, cigarette smoking, alcohol drinking, physical activities, waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and high-density lipoprotein cholesterol. A positive trend was observed between increasing HOMA-IR and prevalence of CKD (P for trend b 0.001) (Table 2). Table 3 showed the crude and multivariate-adjusted ORs of CKD associated with HOMA-IR in participants without diabetes and participants without hypertension. These results were consistent with analyses in general population. However, in participants without MS the multivariate-adjusted ORs of CKD associated with HOMA-IR showed no significance in the second (OR: 1.183, 95% CI: 0.838– 1.670), third (OR: 1.543, 95% CI: 1.103–2.158) and fourth (OR: 1.549, 95% CI: 1.079–2.223) quartile of HOMA-IR relative to the lowest quartile. In addition, sensitivity analyses showed that the primary findings were not influenced in tertile analyses and the results were consistent (data not shown).
4. Discussion In this cross-sectional study of a large nationwide cohort of Chinese population, we provided the evidence that IR was significantly associated with the presence of CKD, defined as an estimated GFR b60 ml/min/1.73 m2, in general population. However, the association lost its significance after excluding patients with MS, indicating that MS may contribute greatly to the association between IR and CKD, and IR was not an independently significant predictor of CKD in Chinese population. Similar with ours, most previous studies have suggested an association between IR and increased risk of CKD. For example, Chen et al. analyzed the data of the Third National Health and Nutrition Examination Survey which included 6453 US adults without diabetes. The results showed that the prevalence of CKD, defined as an estimated GFR b60 ml/min/1.73 m2, was significantly higher with increasing levels of HOMA-IR, and the odds ratio of CKD for the highest compared with the lowest quartile was 2.65 (95% CI: 1.25–5.62, P = 0.008) [15]. In the Health, Aging and Body Composition study which included older individuals without diabetes, eGFR (odds ratio per 10 ml/min/1.73 m2: 0.92, 95% CI: 0.87–0.98) and CKD (1.41, 95% CI: 1.04–1.92) were both independently associated with HOMA-IR [13]. Similarly, Mykkanen et al. described a strict and independent association between IR, estimated by a frequently sampled intravenous glucose tolerance test, and CKD, defined as having microalbuminuria, in non-diabetic individuals [24]. Diabetes has been identified as a major underlying cause of kidney dysfunction [25]. Several cohort studies have indicated an association between diabetes and the increased risk of end-stage renal disease [26,27]. Therefore, most studies focusing the relationship between IR and CKD excluded individuals with diabetes. However, any insulin resistance-related abnormalities may contribute to kidney dysfunction
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
235
Fig. 1. Prevalence of chronic kidney disease in general population (A), population without diabetes (B), population without hypertension (C), and population without metabolic syndrome (D) according to the quartiles of HOMA-IR. HOMA-IR: homeostasis model assessment of insulin resistance; chronic kidney Disease (CKD) was defined as an estimated GFR b60 ml/min/1.73 m2.
[28]. Each single component of constellation of abnormalities clustering under the definition of metabolic syndrome may in fact play a role in kidney damage [28]. For example, abnormal blood pressure may lead to CKD mainly due to intraglomerular hypertension [29]. Generally speaking, when evaluating the effect of one factor to an association, the best way is subgroup analysis, rather than multivariable analysis. Therefore, in order to investigate the influence of MS in the association between IR and CKD, a subgroup analysis by excluding population with MS was conducted. It is known that the group of patients without MS could have 0–2 characteristics of MS and all five characteristics contribute with a different weight to CKD prevalence. So in the subgroup analyses, all the related metabolic variables (e.g. waist circumference, blood glucose, blood pressure) were still adjusted for. However, in the subgroup analysis which excluded participants with MS, we found that the association between IR and CKD lost its significance. Our findings were different from those in Jang's study [17]. Jang et al. conducted a cohort study which included 60,047 participants to evaluate the effect of IR on the 2-year development of CKD, defined as having albuminuria. The results showed that 5th quintile of HOMA-IR compared to 1st quintile of HOMA-IR was associated with the
development of albuminuria in subjects without MS (OR: 1.51, 95% CI: 1.09–2.11, P = 0.014). However, the associations were not significant in other quintiles of HOMA-IR compared to the 1st quintile [17]. Similarly, our results were in accord with Park's study [16], which included 17,157 subjects above 20 years old who underwent voluntary health check-ups. The results showed that while HOMA-IR index values were higher in lower eGFR groups compared with higher eGFR groups (P b 0.001), no meaningful differences in HOMA-IR values according to eGFR groups were observed in subjects without MS [16]. Therefore, our study showed that IR is not independently associated with CKD in Chinese population; our study also indicated that MS, as a constellation of metabolic abnormalities, may contribute greatly to and modify the association of IR with CKD, despite individual component may not influence the significance of the association. The major strength of our study was the nationwide populationbased representatives, so the findings could be generalized to Chinese population. However, several limitations should be addressed. Firstly, the cross-sectional study design limited to draw conclusion regarding the causality among IR, MS and CKD. Our results could not answer the questions whether IR and MS contributed to the initiation or
Table 2 Crude and multivariate-adjusted odds ratio of chronic kidney disease associated with HOMA-IR in general population. Variable
n/N % ± SE Model 0 Model 1 Model 2 Model 3
Q1
Q2
OR
OR (95% CI)
72/2780 2.6 ± 0.3 1.000 (ref.) 1.000 (ref.) 1.000 (ref.) 1.000 (ref.)
93/2788 3.3 ± 0.3 1.288 (0.922–1.800) 1.209 (0.861–1.697) 1.165 (0.829–1.639) 1.183 (0.838–1.670)
Q3 P
OR (95% CI)
NS NS NS NS
117/2788 4.2 ± 0.4 1.724 (1.255–2.367) 1.581 (1.146–2.182) 1.519 (1.097–2.102) 1.543 (1.103–2.158)
Q4 P
OR (95% CI)
0.001 0.005 0.012 0.011
133/2787 4.8 ± 0.4 1.933 (1.416–2.639) 1.573 (1.146–2.159) 1.563 (1.135–2.151) 1.549 (1.079–2.223)
P for trend P b0.001 b0.001 0.005 0.006 0.018
HOMA-IR: homeostasis model assessment of insulin resistance. Model 0: unadjusted; Model 1: adjusted for age, gender and ethnics; Model 2: Model 1 plus further adjusted for educational level, yearly family income, cigarette smoking, alcohol drinking, and physical activities; Model 3: Model 2 plus further adjusted for waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and HDL cholesterol.
236
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237
Table 3 Crude and multivariate-adjusted odds ratio of chronic kidney disease associated with HOMA-IR in population without metabolic syndrome. Variable
Q1
Q2
OR
OR (95% CI)
Participants without diabetes n/N 70/2728 % ± SE 2.6 ± 0.3 Crude OR 1.000 (ref.) Adjusted OR 1.000 (ref.)
89/2672 3.3 ± 0.3 1.286 (0.913–1.811) 1.226 (0.863–1.734)
Participants without hypertension n/N 44/2205 % ± SE 2.0 ± 0.3 Crude OR 1.000 (ref.) Adjusted OR 1.000 (ref.)
56/2100 2.7 ± 0.3 1.291 (0.848–1.966) 1.212 (0.788–1.864)
Participants without metabolic syndrome n/N 56/2458 65/2252 % ± SE 2.3 ± 0.3 2.9 ± 0.3 Crude OR 1.000 (ref.) 1.214 (0.829–1.776) Adjusted OR 1.000 (ref.) 1.178 (0.794–1.747)
Q3
Q4
P
OR (95% CI)
NS NS
105/1927 4.1 ± 0.3 1.708 (1.232–2.369) 1.622 (1.151–2.286)
NS NS
60/1927 3.1 ± 0.3 1.587 (1.052–2.393) 1.485 (0.972–2.271)
NS NS
59/1927 3.1 ± 0.3 1.382 (0.941–2.030) 1.366 (0.909–2.052)
P
OR (95% CI)
0.001 0.006
85/2128 4.0 ± 0.4 1.579 (1.119–2.229) 1.459 (1.005–2.120)
0.028 NS
66/1591 4.1 ± 0.4 2.000 (1.329–3.009) 1.717 (1.112–2.651)
NS NS
40/1244 3.2 ± 0.4 1.256 (0.806–1.957) 1.314 (0.803–2.151)
P for trend P 0.002 0.009 0.047 b0.001 0.001 0.015
NS NS NS
HOMA-IR: homeostasis model assessment of insulin resistance. Adjusted OR: adjusted for age, gender, ethnics, educational level, yearly family income, cigarette smoking, alcohol drinking, physical activities, waist circumference, fasting blood glucose, systolic blood pressure, diastolic blood pressure, serum triglyceride and HDL cholesterol.
progression of CKD, or vice versa. Secondly, GFR was not directly measured, since the methodologies are complex and not feasible in a large epidemiologic study. We thus adopted a modified MDRD equation to assess GFR which showed lower bias and the higher accuracy in each stage of CKD in Chinese population [20]. Similarly, although the euglycemic clamping method is the gold standard for measuring IR, the technique is not feasible in a large epidemiologic study and HOMA-IR was thus used. Thirdly, although we had a relative large study sample, we noticed that the sample size of patients with CKD in the different quartiles were low. The study may not be powered to draw strong conclusions due to low power to demonstrate a difference. Fourthly, we did not have the data of albuminuria and the frequency of CKD was underestimated. The power of the findings would be more adequate if having the data of albuminuria. Finally, a large amount of participants were excluded in the present study, which would be expected to bias the study results to some extent. However, most of the participants were excluded based on centers; for individual included center, the study was well-designed and the response rate was estimated above 80%. In summary, by using a nationwide representative sample, we indicated that IR was not an independently significant predictor of CKD in Chinese population, and MS may contribute greatly to the association between IR and CKD. More studies, particular cohort studies, are required to confirm the present findings. Acknowledgment We thank all physicians and participants of the study, for their cooperation and generous participation. This study was supported by the Chinese Medical Association Foundation and the Chinese Diabetes Society. QJ was supported by the Xijing Hospital Foundation (Grant No. XJZT13Z04). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cca.2015.07.013. References [1] L. Zhang, F. Wang, L. Wang, et al., Prevalence of chronic kidney disease in China: a cross-sectional survey, Lancet 379 (2012) 815–822. [2] R.A. Nugent, S.F. Fathima, A.B. Feigl, D. Chyung, The burden of chronic kidney disease on developing nations: a 21st century challenge in global health, Nephron Clin. Pract. 118 (2011) c269–277.
[3] M.T. Liao, C.C. Sung, K.C. Hung, et al., Insulin resistance in patients with chronic kidney disease, J. Biomed. Biotechnol. 2012 (2012) 691369. [4] G. Bodlaj, J. Berg, R. Pichler, G. Biesenbach, Prevalence, severity and predictors of HOMA-estimated insulin resistance in diabetic and nondiabetic patients with endstage renal disease, J. Nephrol. 19 (2006) 607–612. [5] B. Satirapoj, O. Supasyndh, P. Phantana-Angkul, et al., Insulin resistance in dialysis versus non dialysis end stage renal disease patients without diabetes, J. Med. Assoc. Thai. 94 (Suppl. 4) (2011) S87–93. [6] K.G. Alberti, R.H. Eckel, S.M. Grundy, et al., Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity, Circulation 120 (2009) 1640–1645. [7] S.M. Grundy, J.I. Cleeman, S.R. Daniels, et al., Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement, Circulation 112 (2005) 2735–2752. [8] J. Ming, S. Xu, C. Yang, et al., Metabolic syndrome and chronic kidney disease in general Chinese adults: results from the 2007–08 China National Diabetes and Metabolic Disorders Study, Clin. Chim. Acta 430 (2014) 115–120. [9] B.R. Johns, A.C. Pao, S.H. Kim, Metabolic syndrome, insulin resistance and kidney function in non-diabetic individuals, Nephrol. Dial. Transplant. 27 (2012) 1410–1415. [10] T. Basturk, A. Unsal, Is insulin resistance a risk factor for the progression of chronic kidney disease? Kidney Blood Press. Res. 34 (2011) 111–115. [11] B. Satirapoj, O. Supasyndh, A. Boonyavarakul, L. Luesutthiviboon, P. Choovichian, The correlation of insulin resistance and renal function in non diabetic chronic kidney disease patients, J. Med. Assoc. Thail. 88 (Suppl. 3) (2005) S97–104. [12] H. Pham, C. Robinson-Cohen, M.L. Biggs, et al., Chronic kidney disease, insulin resistance, and incident diabetes in older adults, Clin. J. Am. Soc. Nephrol. 7 (2012) 588–594. [13] M. Landau, M. Kurella-Tamura, M.G. Shlipak, et al., Correlates of insulin resistance in older individuals with and without kidney disease, Nephrol. Dial. Transplant. 26 (2011) 2814–2819. [14] S. Kobayashi, K. Maesato, H. Moriya, T. Ohtake, T. Ikeda, Insulin resistance in patients with chronic kidney disease, Am. J. Kidney Dis. 45 (2005) 275–280. [15] J. Chen, P. Muntner, L.L. Hamm, et al., Insulin resistance and risk of chronic kidney disease in nondiabetic US adults, J. Am. Soc. Nephrol. 14 (2003) 469–477. [16] J.H. Park, S.W. Oh, S.Y. Ahn, et al., Decreased estimated glomerular filtration rate is not directly related to increased insulin resistance, Diabetes Res. Clin. Pract. 99 (2013) 366–371. [17] C.M. Jang, Y.Y. Hyun, K.B. Lee, H. Kim, Insulin resistance is associated with the development of albuminuria in Korean subjects without diabetes, Endocrine 48 (2015) 203–210. [18] W. Yang, J. Lu, J. Weng, et al., Prevalence of diabetes among men and women in China, N. Engl. J. Med. 362 (2010) 1090–1101. [19] D.R. Matthews, J.P. Hosker, A.S. Rudenski, et al., Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man, Diabetologia 28 (1985) 412–419. [20] Y.C. Ma, L. Zuo, J.H. Chen, et al., Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease, J. Am. Soc. Nephrol. 17 (2006) 2937–2944. [21] K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification, Am. J. Kidney Dis. 39 (2002) S1–266. [22] K.G. Alberti, P.Z. Zimmet, Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes
C. Jing et al. / Clinica Chimica Acta 448 (2015) 232–237 mellitus provisional report of a WHO consultation, Diabet. Med. 15 (1998) 539–553. [23] A.V. Chobanian, G.L. Bakris, H.R. Black, et al., Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure, Hypertension 42 (2003) 1206–1252. [24] L. Mykkanen, D.J. Zaccaro, L.E. Wagenknecht, et al., Microalbuminuria is associated with insulin resistance in nondiabetic subjects: the insulin resistance atherosclerosis study, Diabetes 47 (1998) 793–800. [25] S.B. Ghaderian, F. Hayati, S. Shayanpour, S.S. Beladi Mousavi, Diabetes and end-stage renal disease; a review article on new concepts, J. Renal Inj. Prev. 4 (2015) 28–33.
237
[26] W.F. Keane, B.M. Brenner, D. de Zeeuw, et al., The risk of developing end-stage renal disease in patients with type 2 diabetes and nephropathy: the RENAAL study, Kidney Int. 63 (2003) 1499–1507. [27] F.L. Brancati, P.K. Whelton, B.L. Randall, et al., Risk of end-stage renal disease in diabetes mellitus: a prospective cohort study of men screened for MRFIT. Multiple Risk Factor Intervention Trial, JAMA 278 (1997) 2069–2074. [28] Y.J. Sheen, W.H. Sheu, Metabolic syndrome and renal injury, Cardiol. Res. Pract. 2011 (2011) 567389. [29] M.E. Hall, J.M. do Carmo, A.A. da Silva, et al., Obesity, hypertension, and chronic kidney disease, Int. J. Nephrol. Renov. Dis. 7 (2014) 75–88.
