Abstract
Aims/hypothesis
To examine the association of serum cystatin C with the incidence of type 2 diabetes mellitus over a 15-year follow-up period.
Methods
The 15-year cumulative incidence of diabetes was measured in a cohort of Beaver Dam Eye Study participants (n=3472, 1988–2003). A person was defined as developing diabetes (a positive history of diabetes mellitus treated with insulin, oral hypoglycemic agents and/or diet, or had elevations in glycosylated hemoglobin levels) in the absence of diabetes at baseline. The relation of cystatin C and other risk factors to incident type 2 diabetes was determined using discrete time extension of the proportional hazards model.
Results
The 15-year cumulative incidence of diabetes was estimated to be 9.6%. After controlling for age, sex, body mass index, smoking status, glycosylated hemoglobin, proteinuria, chronic kidney disease status, and hypertension status, serum cystatin C at baseline was associated with the 15-year cumulative incidence of type 2 diabetes (Odds Ratio per log of cystatin C unit 2.19, and 95% Confidence Interval 1.02, 4.68).
Conclusions/interpretation
These findings show a positive relationship of serum cystatin C levels with the incidence of type 2 diabetes mellitus independently of confounding risk factors. The findings strongly suggest the need for further evaluation of the potential importance of cystatin C in pathogenesis of type 2 diabetes mellitus.
Keywords: Epidemiology, Prediction and prevention of type 2 diabetes, Pathogenic mechanisms
INTRODUCTION
Type 2 diabetes mellitus is common and an important cause of morbidity and mortality. Its incidence is expected to increase by two to three times over the next 30 to 40 years due largely to changes in diet and physical activity levels [1,2]. Other factors such as smoking and renal disease have also been shown to increase the risk of type 2 diabetes mellitus. Recent studies suggest that cystatin C is not only a sensitive marker of renal dysfunction but it is also associated with insulin resistance, obesity, and hypertension [3,4]. These conditions are known to be related to the development of type 2 diabetes [5,6].
Serum cystatin C has been suggested as being associated with the risk of progression from normoglycemia to pre-diabetes in Western New York Health Study participants [3]. To our knowledge, there are no population based studies that have examined the association of serum cystatin C with the long-term incidence of type 2 diabetes mellitus. In this paper, we examine this relationship in a cohort of participants in the Beaver Dam Eye Study (BDES).
RESEARCH DESIGN AND METHODS
Population
Methods used to identify and describe the population have appeared in previous reports [7–10]. In brief, a private census of the population of Beaver Dam, Wisconsin, was performed from September 15, 1987, to May 4, 1988, to identify all residents in the city or township of Beaver Dam who were 43 to 84 years of age. Of the 5,924 eligible individuals, 4,926 participated in the baseline examination between March 1, 1988, and September 14, 1990 [7]. Ninety-nine percent of the population was white. Of these, 3,684 (81.1%) participated in the 5-year follow-up examination in 1993 to 1995 [8]. Of the 3,334 surviving participants in the baseline and second examination, 2,764 (82.9%) participated in the 10-year follow-up examination [9]. Of the 2,480 surviving participants who were examined at the baseline, 5 and 10-year follow-up examinations, 2,119 (85.4%) participated in the 15-year follow-up examination [10]. Comparisons between participants and non-participants at the time of the baseline, 5, 10, and 15-year follow-up examinations have appeared elsewhere [7–10].
Procedures and Definitions
Similar procedures, used at both the baseline and follow-up examinations, have been described in detail elsewhere [11,12]. Informed consent was obtained from each participant at the beginning of the examination, and all examinations followed a similar protocol that was approved by the Institutional Review Board of the University of Wisconsin and carried out in accordance with the Declaration of Helsinki.
Casual blood specimens and non-fasting blood specimens were obtained from participants. Serum glucose level was determined using the hexokinase method. During the first two exam phases, HbA1c was determined using affinity chromatography (Isolab Inc, Akron, OH, USA). During the third exam phase, total glycosylated hemoglobin was measured using ion capture method (Abbott IMX). During the fourth exam phase, glycosylated hemoglobin A1C was measured using high performance liquid chromatography (Tosoh A1c 2.2 Plus). Total glycosylated hemoglobin from the first three exam phases was standardized to the percentage of glycosylated hemoglobin A1C using the formula: glycosylated hemoglobin A1C=(glycosylated hemoglobin + 1.76)/1.49. Diabetes was defined as present regardless of HbA1c level if a person had been told by a physician that he/she had diabetes mellitus and was treated with insulin, oral hypoglycemic agents and/or diet. In persons diagnosed to have diabetes by a physician but not on treatment, diabetes was defined as present if HbA1c was equal to or greater than 6.5%. Primary care physicians were consulted whenever there was question about past diagnosis. Newly identified diabetes was defined as having an HbA1c level equal to or greater than 7.0% in persons who had not been previously told by a physician that he/she had diabetes. A person was defined as suspected of having diabetes if he/she had a positive history of diabetes, but was not treated with medications and/or diet and glycosylated hemoglobin levels A1C were less than 6.5%. Cumulative incidence of type 2 diabetes mellitus was defined as it being present at any follow-up examination in persons in whom it was absent at the 1988–1990 baseline examination. Persons who were suspected of having diabetes were excluded for these analyses.
Remaining serum was stored without preservative at −80°C in cryogenic vials with Orings for up to 17 years until the vials were shipped on dry ice to the University of Minnesota laboratory for the analyses reported herein. Creatinine was measured in serum by reflectance spectrophotometry on the Vitros analyzer (Johnson & Johnson Clinical Diagnostics, Inc., Rochester, NY, USA). The laboratory CV was 2.2%. The level of high sensitivity C-reactive protein (hsCRP) was measured on the Hitachi 911 (Roche Diagnostics, Indianapolis, IN, USA) using the CRP K-Assayc, a particle enhanced immunonephelometric assay (Kamiya Biomedical Company, Seattle, WA, USA). The expected normal values are ≤5 mg/l and the laboratory CV ranged from 2.1–4.5%. Serum cystatin C was determined nephelometrically using the Dade Behring BN100 nephelometer (Deerfield, IL, USA). The inter-assay precision was determined at two control levels: 1.72 mg/l (CV 6.4%) and 0.78 mg/l (CV 5.2%). GFR was estimated based on the 4-variable Modification of Diet in Renal Disease (MDRD) Study Equation (GFR = 175 ? serum creatinine−1.154 ? age−0.203 ? 1.212 [if black] ? 0.742 [if female]) [13]. Additionally, GFR was calculated based on serum cystatin C levels using the formula (GFR = 127.7 ? serum cystatin C−1.17 ? age−0.13 ? 1.06 [if black] ? 0.91 [if female]) [14]. Gross proteinuria was determined by dipstick on a casual urine sample and defined as having urinary protein concentrations of greater than 0.3 g/l.
Age was defined as the age at the time of the 1988–1990 examination. Cigarette smoking status was determined as follows: a subject was classified as a non-smoker if he/she had smoked fewer than 100 cigarettes in his/her lifetime; as an ex-smoker if he/she had smoked 100 or more cigarettes in his/her lifetime but had stopped smoking before the examination; and as a current smoker if he/she had not stopped smoking. Hypertension was defined as a mean systolic blood pressure of ≥140 mmHg and/or a mean diastolic blood pressure of ≥90 mmHg and/or a history of hypertension using antihypertensive medication at the time of the baseline examination. BMI was defined as weight in kg divided by height in m2. A person was defined as having a history of cardiovascular disease at the 1988–1990 examination based on the participant’s history of angina, myocardial infarction, or stroke. A person was defined as having a sedentary lifestyle based on a history of exercise less than 3 times per week.
Statistical Methods
The cumulative 15-year incidence of diabetes mellitus was estimated accounting for the competing risk of death [15]. Associations between serum cystatin C and other risk factors with time to the development of incident diabetes were assessed using the discrete time extension of the proportional hazards model [16] as implemented in PROC PHREG procedure (SAS Institute, Cary, North Carolina). Effects of covariates on the discrete hazard function are summarized as odds ratios (OR) and 95 percent CI as implemented in PROC PHREG procedure (SAS Institute, Cary, North Carolina). These analytical approaches allowed those people who were not seen after the 5, 10 or 15-year examination due to death or nonparticipation to contribute information to the estimates [15,17]. The assumption of proportionality (constant OR over time) was assessed graphically for all covariates [18]. Analyses were stratified based on baseline smoking status due to non-proportionality of the smoking effect on incident diabetes. Potential interactions between cystatin C and other risk factors were assessed. None were significant at 5% level.
We considered age, sex, BMI, smoking status, HbA1c, hsCRP, chronic kidney disease status, hypertension, sedentary lifestyle, and gross proteinuria history as potential confounders in multivariate analysis. The distribution of the serum cystatin C was highly skewed and serum cystatin C was not linearly associated with the outcome variable, so we performed logarithmic transformation of cystatin C to use in the analysis.
RESULTS
Of the 4,924 people who participated at the baseline examination, 4,901 had information on diabetes status; 441 (9.0%) had diabetes and 92 (1.9%) were suspected of having diabetes. Of the remaining 4,368 subjects without diabetes, 896 (18.3%) were alive but not examined at the 5-year follow-up examination due to refusal or they were unable to be located. Incidence analyses in this report are limited to the remaining 3,472 people at risk of developing diabetes, including 714 people who died at any follow-up examination. Persons without diabetes incidence data were more likely to be older and, after adjusting for age, were more likely to be a current smoker and have a higher mean serum cystatin C than participants included in analyses.
The 15-year cumulative incidence of diabetes was estimated to be 9.6%. Characteristics of the study population by those who did and did not develop diabetes or died are presented in Table 1. While adjusting for age and sex, people with higher levels of serum cystatin C were more likely to die (p<0.05). Also, people who died had more frequent history of cardiovascular disease and gross proteinuria history (p<0.05). People who developed diabetes during the 15-year follow-up were older, had higher BMI, and were more likely to be male and a current smoker (p<0.05). Additionally, people who developed diabetes had higher mean glycosylated hemoglobin levels and mean systolic and diastolic BP at the baseline examination, but did not differ by mean serum total cholesterol levels from people who did not develop diabetes (p<0.05). Also, people who developed diabetes were more likely to be hypertensive, were more likely to have a history of cardiovascular disease and to use beta-blockers, diuretics and nonsteroidal anti-inflammatory drugs (NSAIDs) than people who did not develop diabetes. People who developed diabetes had higher serum cystatin C levels and lower cystatin C based GFR, but did not differ by serum creatinine and creatinine based GFR levels and gross proteinuria status from people who did not develop diabetes.
Table 1.
Baseline Characteristics by Diabetes Incidence or Death Status at the 1988–1990 Beaver Dam Eye Study Examination.
| Characteristic | Did not Develop T2DM (N=2,483) | Developed T2DM (N=275) | Died (N=714) |
|---|---|---|---|
| Age (yrs), mean ± SD | 58.6 ± 10.0 | 60.0 ± 10.4 | 70.0 ± 9.8 |
| Males, N (%) | 1025 (41.3) | 146 (53.1) | 375 (52.5) |
| BMI, N (%) | |||
| 25–29 kg/m2 | 1077 (43.4) | 76 (27.6) | 275 (38.5) |
| ≥30 kg/m2 | 753 (30.3) | 174 (63.3) | 228 (31.9) |
| Smoking history, N (%) | |||
| Ex-smoker | 827 (33.3) | 124 (45.1) | 284 (39.8) |
| Current smoker | 477 (19.2) | 46 (16.7) | 147 (20.6) |
| Sedentary lifestyle, N (%) | 1829 (73.7) | 205 (74.6) | 575 (80.5) |
| Glycosylated hemoglobin A1c (%), mean ± SD | 5.6 ± 0.6 | 6.2 ± 0.8 | 5.8 ± 0.7 |
| Serum total cholesterol (mg/dl), mean ± SD | 232.5 ± 42.4 | 235.6 ± 49.7 | 230.7 ± 46.4 |
| Serum total cholesterol (mmol/l), mean ± SD | 6.02 ± 1.10 | 6.10 ± 1.29 | 5.98 ± 1.20 |
| Systolic BP (mmHg), mean ± SD | 128.4 ± 18.5 | 135.0 ± 19.0 | 136.1 ± 22.4 |
| Diastolic BP (mmHg), mean ± SD | 77.9 ± 10.1 | 79.7 ± 11.3 | 75.2 ± 11.7 |
| Presence of hypertension, N (%) | 1037 (41.8) | 169 (61.5) | 428 (60.1) |
| Cardiovascular disease history, N (%) | 196 (8.0) | 40 (14.6) | 193 (27.6) |
| History of beta-blocker use, N (%) | 160 (6.5) | 33 (12.0) | 72 (10.1) |
| History of calcium channel blocker use, N (%) | 39 (1.6) | 7 (2.6) | 20 (2.8) |
| History of diuretic use, N (%) | 471 (19.2) | 80 (29.9) | 266 (38.0) |
| History of ACE inhibitors use, N (%) | 111 (4.5) | 18 (6.7) | 58 (8.2) |
| History of NSAID use, N (%) | 769 (31.0) | 100 (36.4) | 271 (38.0) |
| Presence of gross proteinuria, N (%) | 52 (2.2) | 12 (4.6) | 46 (7.0) |
| Serum hsCRP (mg/l), median (interquartile range) | 1.67 (0.8–3.4) | 2.65 (1.3–4.8) | 2.77 (1.3–6.4) |
| Serum hsCRP (nmol/l), median (interquartile range) | 15.9 (7.6–32.3) | 25.2 (12.4–45.6) | 26.3 (12.4–60.8) |
| Serum creatinine (mg/dl), mean ± SD | 0.89 ± 0.19 | 0.93 ± 0.23 | 1.01 ± 0.48 |
| Serum creatinine (μmol/l), mean ± SD μmol/l | 78.7 ± 16.8 | 82.2 ± 20.3 | 89.28 ± 42.4 |
| History of CKDa, N (%) | 272 (11.5) | 34 (12.1) | 189 (29.1) |
| GFRb (ml min−1 1.73 m−2), mean ± SD | 78.1 ± 16.2 | 77.6 ± 17.6 | 72.1 ± 22.3 |
| GFRc (ml min−1 1.73 m−2), mean ± SD | 91.8 ± 21.9 | 82.8 ± 19.7 | 69.5 ± 26.0 |
| Serum cystatin C (mg/l), mean ± SD | 0.86 ± 0.19 | 0.93 ± 0.22 | 1.09 ± 0.43 |
| Serum cystatin C (nmol/l), mean ± SD | 64.4± 14.2 | 70.0± 16.5 | 81.6± 32.2 |
Abbreviations: CKD=chronic kidney disease; NSAID=non-steroidal anti-inflammatory drugs; hsCRP=high sensitivity-C-reactive protein; T2DM=type 2 diabetes mellitus.
CKD status was defined using serum creatinine by GFR<60 ml min−1 1.73 m−2.
GFR is based on serum creatinine.
GFR is based on serum cystatin C.
Our median value and interquartile range for cystatin C were 0.85 and 0.75–0.99, correspondingly. On the log scale it corresponded to median value of −0.16 and interquartile range of −0.29 to −0.01. The geometric mean of cystatin C was 0.83 for people who did not develop diabetes, 0.90 for those who developed diabetes and 1.03 for those who died.
The results of multivariate analysis showed that serum cystatin C at baseline was associated with diabetes development after adjustment for sex and age (Table 2). Further sequential adjustment for BMI, smoking status, glycosylated hemoglobin, hypertension history, gross proteinuria and chronic kidney disease status showed that serum cystatin C remained associated with the incidence of diabetes (OR per log of cystatin C 2.19; 95% CI 1.02, 4.68). It corresponded to OR 1.15; 95% CI 1.01; 1.31 per 50% increase of cystatin C. Serum cystatin C in our study was weakly correlated with markers of inflammation such as hsCRP (r=0.19, p=0.0001). Meanwhile, the association of serum cystatin C with the incidence of diabetes remained significant after further adjustment for serum C-reactive protein (OR 2.09; 95% CI 1.02, 4.11). Estimated GFR based on serum cystatin C levels was related to the incidence of diabetes after controlling for age and sex (OR ml min−1 1.73 m−2 0.98, 95% CI 0.97, 0.99, p<0.0001). In contrast, GFR based on serum creatinine using the MDRD equation was not related to the incidence of diabetes after controlling for age and sex (OR ml min−1 1.73 m−2 1.00, 95%CI 0.99, 1.01, p=0.94). Stratification by kidney disease status (Table 2) showed that the association of serum cystatin C with the incidence of diabetes was similar in those people with and without CKD (Table 2). The association of serum cystatin C with the incidence of diabetes was significant after adjustment for age and sex in both groups. Body mass index was the strongest confounder of the association of cystatin C with incident diabetes in both groups. The association did not remain significant after adjustment for BMI in persons with kidney disease and after the adjustment for glycosylated hemoglobin in persons without kidney disease. After controlling for confounding factors, the odds of having incident diabetes was not different in both subgroups (Table 2).
Table 2.
Impact of Sequential Adjustment on Cystatin C Related Differences in 15-year Cumulative Diabetes Incidence in the Beaver Dam Eye Study.
| Variables | Overall (N=3,472) | No CKD, no proteinuria (N=2,706) | CKD or proteinuria (N=571) |
|---|---|---|---|
| OR per log of cystatin C (95% CI) | OR per log of cystatin C (95% CI) | OR per log of cystatin C (95% CI) | |
| Adjusted for agea and sex | 4.62 (2.52, 8.49) | 6.74 (2.78; 16.35) | 4.34 (1.40;13.43) |
| Adjusted for all above and BMI | 2.51 (1.32, 4.79) | 2.76 (1.11; 6.87) | 2.83 (0.89; 9.04) |
| Adjusted for all above and smoking status | 2.50 (1.30, 4.82) | 2.52 (1.01; 6.36) | 3.28 (0.96, 1.02) |
| Adjusted for all above and glycosylated hemoglobin | 2.24 (1.16, 4.33) | 2.40 (0.91; 6.30) | 2.54 (0.74; 8.69) |
| Adjusted for all above and hypertension history | 1.99 (1.03, 3.85) | 2.06 (0.79; 5.43) | 2.26 (0.65; 7.91) |
| Adjusted for all above and gross proteinuria and CKD statusb | 2.19 (1.02, 4.68) | - | - |
Abbreviations: CKD=chronic kidney disease.
Age is entered in analysis as a continuous variable. Logarithmic transformation of continuous variable was performed for serum cystatin C.
CKD status was defined using serum creatinine by GFR<60 ml min−1 1.73 m−2.
The difference between overall group and stratified groups in table 2 is explained by missing values for proteinuria and/or CKD status in these groups
DISCUSSION
In our study, while controlling for age, sex, BMI, hypertension status, glycosylated hemoglobin and other risk factors, serum level of cystatin C was associated with increased risk of incident type 2 diabetes mellitus. To our knowledge, this is the first long-term population-based study that shows an association of serum cystatin C with the incidence of type 2 diabetes mellitus. While we found the effect was independent of the presence of chronic kidney disease as defined by the MDRD equation and proteinuria, the association may still be due to kidney disease as suggested by the association of the estimated glomerular filtration rate using a cystatin C equation with increased risk of diabetes in our study. Serum cystatin C is thought to be more precise than serum creatinine in defining the presence of early kidney dysfunction [19–21]. It is possible that the preclinical stage of kidney dysfunction, defined by a higher serum cystatin C than estimated by the MDRD study equation, might explain its relationship to type 2 diabetes incidence in our study. Kidney dysfunction defined by low estimated GFR has been shown by others to be related to the development of diabetes [22]. This has, in part, thought to be due to insulin resistance [23], increased renal gluconeogenesis [24], endothelial dysfunction and/or chronic inflammation [25,26], oxidative stress [26], and activation of the renin-angiotensin system [27] in persons with chronic kidney disease.
It is also possible that the relationship of serum cystatin C to the incidence of diabetes is the result of other disease processes. In persons without kidney disease, the association of serum cystatin C with incident diabetes was significant after controlling for age, sex, BMI, and smoking status. However, the association did not remain significant after further adjustment for glycosylated hemoglobin. Recent studies suggest that cystatin C is associated with insulin resistance, obesity and hypertension, conditions closely related to diabetes [3,4]. Additionally, cystatin C is related to inflammation and oxidative stress, which are key pathogenetic components of the metabolic syndrome [3]. Oxidative stress has been shown to induce the synthesis of cys-C mRNA and protein, which may reflect a cellular defensive response to oxidative stress [28]. Higher serum cystatin C values have been shown to be directly related to chronic inflammation [29]. Some studies have linked chronic inflammation to diabetes incidence [30,31]. The relationship of higher serum levels of cystatin C to diabetes development in our study might be partially explained by a direct contribution of serum cystatin C to the interrelated processes of inflammation and endothelial dysfunction, thus increasing the risk of developing diabetes. Our secondary analyses showed that hsCRP attenuated the association between serum cystatin C and diabetes incidence; however, the relation remained statistically significant.
While our study has many strengths, including the long length of follow-up, high participation rate, and the use of standardized protocols to measure risk factors and diabetes, it also has some limitations. First, the small number of nonwhite and Hispanic participants may limit the ability to generalize our findings to other ethnic groups. Second, it is possible that selective survival (persons with higher levels of cystatin C and other risk factors who developed diabetes were less likely to be seen in follow-up due to higher mortality) resulted in underestimation of the relationship between cystatin C and diabetes. Third, there may be uncontrolled confounding by unmeasured characteristics such as genetic or nutritional factors that may be related to cystatin C and diabetes. Fourth, there may be misclassification of diabetes in our study. However, we checked our results eliminating persons with borderline HbA1C levels (greater and equal to 6.0% but less and equal to 6.5%); our results yielded the same conclusion.
In summary, serum cystatin C levels were associated with incident type 2 diabetes independently of confounding risk factors. The exact mechanism is unknown but our results suggest that cystatin C might serve as a more specific marker than serum creatinine for early stages of chronic kidney disease, a risk factor for the incidence of diabetes or it may contribute directly to the pathogenesis of type 2 diabetes mellitus.
Acknowledgments
This project described was supported by EY06594 from the National Eye Institute and DK073217 from the National Institute of Diabetes and Digestive and Kidney Diseases, both from the National Institute of Health, Bethesda, MD, as well as by a Mentor-Based Postdoctoral Fellowship Award from the American Diabetes Association, Alexandria, VA. The content is solely the responsibility of the authors and does not necessarily reflect the official views of the National Eye Institute, the National Institute of Diabetes and Digestive and Kidney Diseases, nor the National Institutes of Health.
The authors would like to acknowledge Ronald E. Gangnon of the University of Wisconsin School of Medicine and Public Health, Population Health Sciences, for his statistical input.
Abbreviations
- BDES
Beaver Dam Eye Study
- CKD
chronic kidney disease
- hsCRP
high sensitivity C-reactive protein
- MDRD
Modification of Diet in Renal Disease
- NSAIDs
nonsteroidal anti-inflammatory drugs
- T2DM
Type 2 diabetes mellitus
Footnotes
DUALITY OF INTEREST
The authors declare that there is no duality of interest associated with this manuscript.
References
- 1.Geiss LS, Pan L, Cadwell B, Gregg EW, Benjamin SM, Engelgau MM. Changes in incidence of diabetes in U.S. adults, 1997–2003. Am J Prev Med. 2006;30:371–377. doi: 10.1016/j.amepre.2005.12.009. [DOI] [PubMed] [Google Scholar]
- 2.Kirtland KA, Li YF, Geiss LS, Thompson TJ. State specific incidence of diabetes among adults-participating states, 1995–1997, and 2005–2007. 2008;2008:4–9. 2009. http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5743a2.htm.
- 3.Surendar J, Indulekha K, Aravindhan V, Ganesan A, Mohan V. Association of Cystatin-C with Metabolic Syndrome in Normal Glucose-Tolerant Subjects (CURES-97) Diabetes Technol Ther. 2010;12:907–912. doi: 10.1089/dia.2010.0077. [DOI] [PubMed] [Google Scholar]
- 4.Servais A, Giral P, Bernard M, Bruckert E, Deray G, Isnard BC. Is serum cystatin-C a reliable marker for metabolic syndrome? Am J Med. 2008;121:426–432. doi: 10.1016/j.amjmed.2008.01.040. [DOI] [PubMed] [Google Scholar]
- 5.Kahn HS, Cheng YJ, Thompson TJ, Imperatore G, Gregg EW. Two risk-scoring systems for predicting incident diabetes mellitus in U.S. adults age 45 to 64 years. Ann Intern Med. 2009;150:741–751. doi: 10.7326/0003-4819-150-11-200906020-00002. [DOI] [PubMed] [Google Scholar]
- 6.Surampudi PN, John-Kalarickal J, Fonseca VA. Emerging concepts in the pathophysiology of type 2 diabetes mellitus. Mt Sinai J Med. 2009;76:216–226. doi: 10.1002/msj.20113. [DOI] [PubMed] [Google Scholar]
- 7.Klein R, Klein BE, Linton KL, De Mets DL. The Beaver Dam Eye Study: visual acuity. Ophthalmology. 1991;98:1310–1315. doi: 10.1016/s0161-6420(91)32137-7. [DOI] [PubMed] [Google Scholar]
- 8.Klein R, Klein BE, Lee KE. Changes in visual acuity in a population. The Beaver Dam Eye Study. Ophthalmology. 1996;103:1169–1178. doi: 10.1016/s0161-6420(96)30526-5. [DOI] [PubMed] [Google Scholar]
- 9.Klein R, Klein BE, Lee KE, Cruickshanks KJ, Chappell RJ. Changes in visual acuity in a population over a 10-year period : The Beaver Dam Eye Study. Ophthalmology. 2001;108:1757–1766. doi: 10.1016/s0161-6420(01)00769-2. [DOI] [PubMed] [Google Scholar]
- 10.Klein R, Klein BE, Lee KE, Cruickshanks KJ, Gangnon RE. Changes in visual acuity in a population over a 15-year period: the Beaver Dam Eye Study. Am J Ophthalmol. 2006;142:539–549. doi: 10.1016/j.ajo.2006.06.015. [DOI] [PubMed] [Google Scholar]
- 11.Klein R, Klein BE. Manual of Operations. US Department of Commerce; Springfield, VA: 1995. The Beaver Dam Eye Study II. NTIS Accession No. PB95-273827. [Google Scholar]
- 12.Klein R, Klein BE. Manual of Operations. US Department of Commerce; Springfield, VA: 1999. The Beaver Dam Eye Study III. NTIS Accession No. PB99-137861. [Google Scholar]
- 13.Levey AS, Coresh J, Greene T, et al. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247–254. doi: 10.7326/0003-4819-145-4-200608150-00004. [DOI] [PubMed] [Google Scholar]
- 14.Stevens LA, Coresh J, Schmid CH, et al. Estimating GFR using serum cystatin C alone and in combination with serum creatinine: a pooled analysis of 3,418 individuals with CKD. Am J Kidney Dis. 2008;51:395–406. doi: 10.1053/j.ajkd.2007.11.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Gooley TA, Leisenring W, Crowley J, Storer BE. Estimation of failure probabilities in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18:695–706. doi: 10.1002/(sici)1097-0258(19990330)18:6<695::aid-sim60>3.0.co;2-o. [DOI] [PubMed] [Google Scholar]
- 16.Cox DR. Regression models and life-tables (with discussion) J Roy Stat Soc Ser B. 1972;34:187–220. [Google Scholar]
- 17.Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457–481. [Google Scholar]
- 18.Kleinbaum DG. Survival analysis: a self-learning text. Springer; 1996. [Google Scholar]
- 19.Coll E, Botey A, Alvarez L, et al. Serum cystatin C as a new marker for noninvasive estimation of glomerular filtration rate and as a marker for early renal impairment. Am J Kidney Dis. 2000;36:29–34. doi: 10.1053/ajkd.2000.8237. [DOI] [PubMed] [Google Scholar]
- 20.Newman DJ, Thakkar H, Edwards RG, et al. Serum cystatin C measured by automated immunoassay: a more sensitive marker of changes in GFR than serum creatinine. Kidney Int. 1995;47:312–318. doi: 10.1038/ki.1995.40. [DOI] [PubMed] [Google Scholar]
- 21.Willems D, Wolff F, Mekhali F, Gillet C. Cystatin C for early detection of renal impairment in diabetes. Clin Biochem. 2009;42:108–110. doi: 10.1016/j.clinbiochem.2008.10.002. [DOI] [PubMed] [Google Scholar]
- 22.Lorenzo C, Nath SD, Hanley AJ, Abboud HE, Gelfond JA, Haffner SM. Risk of type 2 diabetes among individuals with high and low glomerular filtration rates. Diabetologia. 2009;52:1290–1297. doi: 10.1007/s00125-009-1361-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fox CS, Larson MG, Leip EP, Culleton B, Wilson PW, Levy D. Predictors of new-onset kidney disease in a community-based population. JAMA. 2004;291:844–850. doi: 10.1001/jama.291.7.844. [DOI] [PubMed] [Google Scholar]
- 24.Eid A, Bodin S, Ferrier B, et al. Intrinsic gluconeogenesis is enhanced in renal proximal tubules of Zucker diabetic fatty rats. J Am Soc Nephrol. 2006;17:398–405. doi: 10.1681/ASN.2005070742. [DOI] [PubMed] [Google Scholar]
- 25.Meigs JB, Hu FB, Rifai N, Manson JE. Biomarkers of endothelial dysfunction and risk of type 2 diabetes mellitus. JAMA. 2004;291:1978–1986. doi: 10.1001/jama.291.16.1978. [DOI] [PubMed] [Google Scholar]
- 26.Ramos LF, Shintani A, Ikizler TA, Himmelfarb J. Oxidative stress and inflammation are associated with adiposity in moderate to severe CKD. J Am Soc Nephrol. 2008;19:593–599. doi: 10.1681/ASN.2007030355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tikellis C, Wookey PJ, Candido R, Andrikopoulos S, Thomas MC, Cooper ME. Improved islet morphology after blockade of the renin- angiotensin system in the ZDF rat. Diabetes. 2004;53:989–997. doi: 10.2337/diabetes.53.4.989. [DOI] [PubMed] [Google Scholar]
- 28.Demircan N, Gurel A, Armutcu F, Unalacak M, Aktunc E, Atmaca H. The evaluation of serum cystatin C, malondialdehyde, and total antioxidant status in patients with metabolic syndrome. Med Sci Monit. 2008;14:CR97–101. [PubMed] [Google Scholar]
- 29.Taglieri N, Koenig W, Kaski JC. Cystatin C and cardiovascular risk. Clin Chem. 2009;55:1932–1943. doi: 10.1373/clinchem.2009.128397. [DOI] [PubMed] [Google Scholar]
- 30.Hu FB, Meigs JB, Li TY, Rifai N, Manson JE. Inflammatory markers and risk of developing type 2 diabetes in women. Diabetes. 2004;53:693–700. doi: 10.2337/diabetes.53.3.693. [DOI] [PubMed] [Google Scholar]
- 31.Hu G, Jousilahti P, Tuomilehto J, Antikainen R, Sundvall J, Salomaa V. Association of serum C-reactive protein level with sex-specific type 2 diabetes risk: a prospective finnish study. J Clin Endocrinol Metab. 2009;94:2099–2105. doi: 10.1210/jc.2008-2260. [DOI] [PubMed] [Google Scholar]