Glycosylated Hemoglobin and Mortality in Patients with Nondiabetic Chronic Kidney Disease
Vandana Menon*,
Tom Greene,
Arema A. Pereira*,
Xuelei Wang,
Gerald J. Beck,
John W. Kusek,
Allan J. Collins,
Andrew S. Levey* and
Mark J. Sarnak*
* Department of Medicine, Division of Nephrology, Tufts-New England Medical Center, Boston, Massachusetts, Department of Biostatistics and Epidemiology, Cleveland Clinic Foundation, Cleveland, Ohio, National Institutes of Health, Bethesda, Maryland, and Division of Nephrology, Hennepin County Medical Center, Minneapolis, Minnesota
Address correspondence to: Dr. Mark Sarnak, Division of Nephrology, Department of Medicine, 750 Washington Street, NEMC #391, Boston, MA 02111. Phone: 617-636-1182; Fax: 617-636-8329; E-mail: msarnak{at}tufts-nemc.org
Received for publication May 26, 2005.
Accepted for publication August 30, 2005.
In the general population, hyperglycemia in the absence of diabetesmay be associated with increased risk for mortality. Hyperglycemiais prevalent in chronic kidney disease; however, the relationshipbetween glycosylated hemoglobin (HbA1c) as a marker of chronichyperglycemia and outcomes has not been studied in nondiabeticchronic kidney disease. HbA1c was measured at baseline in therandomized cohort of the Modification of Diet in Renal DiseaseStudy (n = 840). Participants with diabetes (n = 43), fastingglucose levels >126 mg/dl (n = 20), or missing HbA1c levels(n = 9) were excluded. Survival status until December 2000 wasobtained from the National Death Index. Death was classifiedas cardiovascular (CVD) when the primary cause was InternationalClassification of Disease, Ninth Revision codes 390 to 459.Cox models were performed to assess the relationship of HbA1cwith all-cause and CVD mortality. Mean (SD) age was 52 (12)years, and mean (SD) GFR was 32 (12) ml/min per 1.73 m2. Eighty-sixpercent of participants were white, and 61% were male. Mean(SD) HbA1c was 5.6% (0.5). A total of 169 (22%) patients died,96 (13%) from CVD. After adjustment for randomization assignmentsand demographic, CVD, and kidney disease factors, HbA1c wasa predictor of all-cause mortality (hazard ratio per 1% increase1.73; 95% confidence interval 1.24 to 2.41; P = 0.001). Therewas a trend toward statistical significance in the relationshipbetween HbA1c and CVD mortality (hazard ratio per 1% increase1.53; 95% confidence interval 0.96 to 2.43; P = 0.07). HbA1cis associated with increased mortality in nondiabetic kidneydisease. Hyperglycemia may be a potential therapeutic targetand HbA1c may be important as a risk stratification tool inthis high-risk population.
Diabetes is an established cardiovascular (CVD) risk factor;however, a growing literature indicates that the relationshipbetween glucose levels and CVD may extend below the thresholdcurrently defined as diabetes. Impairments in glucose metabolism,manifest as hyperglycemia, are associated with poor prognosisin the general population, in the absence of diabetes (1–5).
Chronic kidney disease (CKD) is a growing public health problemof epidemic proportions; currently approximately 9 million peoplein the United States have GFR of <60 ml/min per 1.73 m2 (6).Reduced kidney function is now recognized as a powerful andindependent risk factor for CVD morbidity and mortality (7,8).The relationship between CKD and CVD does not seem to be fullyexplained by traditional CVD risk factors (9,10).
Disorders of glucose homeostasis are common in CKD. Two studiesusing data from the Third National Health and Nutrition ExaminationSurvey found a high prevalence of impaired fasting glucose levels(defined as >110 mg/dl) among individuals with reduced GFRand those with microalbuminuria (11,12). That kidney diseaseis characterized by hyperglycemia raises the possibility thatimpaired glucose metabolism may be an important contributorto the excess CVD risk seen in this population. No prospectivestudies, however, have evaluated the relationship between hyperglycemiaand outcomes among nondiabetic patients with CKD.
Glycosylated hemoglobin (HbA1c), a measure of chronic hyperglycemia,is a sensitive and reliable marker of impaired glucose metabolism(13,14). Some studies have shown HbA1c to be a predictor offuture CVD events among nondiabetic patients in the generalpopulation (15,16), whereas others have found no association(17) or an association only in women (13,18). Using data fromthe randomized cohort of the Modification of Diet in Renal Disease(MDRD) Study, we performed longitudinal analyses to test thehypothesis that HbA1c is an independent predictor of all-causeand CVD mortality in nondiabetic men and women with CKD beforereaching kidney failure.
The MDRD Study, conducted from 1989 to 1993, was a randomized,controlled trial to study the effect of dietary protein restrictionand strict BP control on the progression of kidney disease (19).A total of 585 patients with a baseline GFR of 25 to 55 ml/minper 1.73 m2 were randomized into study A, and 255 patients witha baseline GFR of 13 to 24 ml/min per 1.73 m2 were randomizedto study B. Patients in study A and study B were combined forthese analyses. We excluded participants with diagnosed diabetes(n = 43), those with fasting glucose levels >126 mg/dl (n= 20), and missing HbA1c levels (n = 9). Thus, our effectivestudy sample size was 768. GFR was assessed by the kidney clearanceof 125I-iothalamate. HbA1c was assayed in the central laboratory,in fasting samples obtained during baseline study visits usingHPLC (Bio-Rad Diamat Automated Glycosylated Hemoglobin Analyzer).
Survival status and date and cause of death were ascertainedfrom the National Death Index. A death was ascribed to CVD whenthe primary cause of death was International Classificationof Diseases, Ninth Revision codes 390 to 459 (n = 85) or whenkidney disease was listed as the primary cause of death andCVD was the secondary cause (n = 11). Survival time was definedas time from randomization to death or end of follow-up (December31, 2000). Data collection procedures were approved by the ClevelandClinic and Tufts-New England Medical Center Institutional ReviewBoards.
Statistical Analyses
Summary statistics, according to quartiles of HbA1c, are presentedas percentages for categorical data, mean (±SD) for approximatelynormally distributed continuous variables, and median (interquartilerange) for skewed continuous variables. Differences betweenthe groups were tested using the 2 test, one-way ANOVA, andthe Kruskall-Wallis test as appropriate.
Incidence rates for mortality were calculated for quartilesof HbA1c. Differences in survival between the quartiles of HbA1cwere compared using Kaplan-Meier survival plots. Cox proportionalhazards models were used to evaluate the relationship betweenHbA1c and all-cause and CVD mortality initially without adjustmentand subsequently adjusting for several groups of a priori definedconfounding variables. Studies A and B were combined for theseanalyses; however, the Cox models were stratified by study toallow different baseline hazard rates in the two studies. Model1 adjusted for randomization assignments to protein diets andBP strata, age, gender, and race. Model 2 adjusted for historyof coronary disease as well as CVD risk factors, namely smokingstatus, body mass index, systolic BP, LDL and HDL cholesterol,and C-reactive protein (CRP), in addition to the variables inmodel 1. Model 3 adjusted for variables in model 2 as well asthe kidney disease factors proteinuria and cause of kidney disease.The univariate and fully adjusted Cox models were repeated usingquartiles of HbA1c. The models were repeated replacing HbA1cwith fasting glucose as a continuous variable.
Hazard ratios (HR) are presented per unit (1%) increase in HbA1c,and 95% confidence intervals (CI) were calculated for the HR.Proportional hazards assumptions were tested using log minuslog survival plots and using plots of Schoenfeld residuals versussurvival time.
Sensitivity Analyses and Interactions
Because stratifying by study may not fully adjust for levelof kidney function, we repeated the multivariable Cox modelsadjusting for baseline GFR as a continuous variable. Hematocrit,use of angiotensin-converting enzyme inhibitors, and aspirinmay be potential confounders of the association between HbA1cand mortality. We therefore repeated the final regression modelwith the addition of hematocrit and baseline medication use.
The American Diabetes Association recommends that the goal oftherapy in type 2 diabetes is an HbA1c level of <7% (20).Whereas HbA1c levels of 4 to 6% are considered normal, levels>6.4% have been used as a cutoff indicative of elevated HbA1c(17). We therefore performed additional analyses after excludingpatients with HbA1c > 6.4% (n = 32) to specifically investigatethe association between HbA1c and mortality when HbA1c is belowcurrently defined thresholds of normal (n = 32).
We evaluated an interaction between HbA1c and gender on thebasis of data from previous studies that have suggested thatthe association between HbA1c and mortality (or CVD) is presentonly in women (13,18,21) or is stronger in women than in men(22). We also evaluated an interaction between HbA1c and studyto assess for differential effects according to severity ofkidney disease.
Baseline Characteristics
The study cohort had mean ± SD age of 52 ± 12yr, GFR 32 ± 12 ml/min per 1.73 m2, and HbA1c 5.6 ±0.5%. The sample was predominantly white (86%), 61% were male,and 10% were current smokers. Higher HbA1c was associated witholder age, black race, and a worse CVD risk profile with higherbody mass index and higher levels of systolic BP, total cholesterol,LDL cholesterol, and CRP and higher baseline aspirin use (Table 1).GFR was lower in the higher HbA1c groups; however, therewere no differences in the other kidney disease–relatedfactors.
Table 1. Baseline characteristics by quartiles of HbA1ca
Outcomes
Median follow-up for analyses of survival was 125 mo. All-causemortality was 22% (n = 169), and CVD mortality was 13% (n =96). Incidence rates for all-cause/CVD mortality were 14/9 per1000 person-years in quartile 1, 14/7 per 1000 person-yearsin quartile 2, 27/17 per 1000 person-years in quartile 3, and38/19 per 1000 person-years in quartile 4. Kaplan-Meier curvesdemonstrated differences in survival between quartiles of HbA1c(Figure 1).
Figure 1. Kaplan-Meier survival curves showed that higher glycosylated hemoglobin levels were associated with higher all-cause mortality (log rank test P < 0.0001).
In univariate analysis, a 1% higher HbA1c was associated witha greater than two-fold increase in the risk for all-cause andCVD mortality (Table 2). After adjustment for randomizationassignments, demographic, CVD, and kidney disease factors, a1% increase in HbA1c was associated with a 73% increase in therisk for all-cause mortality. In multivariable analyses, therewas a trend toward statistical significance in the relationshipbetween HbA1c and CVD mortality.
Table 2. Relationship of HbA1c with all-cause and CVD mortalitya
Our prespecified analysis relating mortality hazard to HbA1cassumed a linear model. However, the quartiles analysis of incidencerates and the Kaplan-Meier curves suggests the possibility ofa threshold in risk for the relationship between HbA1c and all-causeand CVD mortality. Therefore, we present the multivariate Coxregression analysis using quartiles of HbA1c; however, it mustbe acknowledged that our sample size is too small to make definitivestatements regarding these thresholds. Table 3 presents theresults of these analyses. In univariate analysis, quartiles3 and 4 of HbA1c were associated with increased risk for all-causeand CVD mortality relative to quartile 1. After adjustment forrandomization assignments, demographic, CVD, and kidney disease–relatedfactors, only quartile 4 (HR, 1.80; 95% CI, 1.07 to 3.02; P= 0.03) of HbA1c was associated with increased risk for all-causemortality relative to quartile 1. There was no difference inHR for CVD mortality for the three highest HbA1c quartiles comparedwith the lowest quartile.
Table 3. Relationship of quartiles of HbA1c with all-cause and CVD Mortality
Fasting glucose levels were associated with all-cause (HR per1-mg/dl increase in glucose, 1.02; 95% CI, 1.01 to 1.03; P <0.001) and CVD (HR, 1.03; 95% CI, 1.01 to 1.04; P = 0.001) mortalityin unadjusted Cox models. However, these associations were attenuatedfor both all-cause (HR, 1.01; 95% CI, 0.99 to 1.02; P = 0.34)and CVD (HR, 1.01; 95% CI, 0.99 to 1.03; P = 0.11) mortalityafter adjustment for previously described covariates.
Sensitivity Analysis
With the inclusion of GFR in the multivariable Cox models, HbA1cremained a significant predictor of all-cause mortality (HR,1.67; 95% CI, 1.19 to 2.34), and the relationship with CVD mortalitywas essentially unchanged (HR, 1.47; 95% CI, 0.92 to 2.35).The addition of hematocrit did not appreciably alter the HRfor HbA1c as a continuous variable in models that examined all-cause(HR, 1.66; 95% CI, 1.18 to 2.33) or CVD (HR, 1.48; 95% CI, 0.93to 2.35) mortality. Similarly, the HR for HbA1c remained relativelyunchanged with the addition of baseline aspirin and angiotensin-convertingenzyme inhibitor use for both all-cause (HR, 1.70; 95% CI, 1.22to 2.39) and CVD mortality (HR, 1.50; 95% CI, 0.92 to 2.37).After exclusion of patients with elevated HbA1c (defined asHbA1c > 6.4%), HbA1c remained a significant predictor ofall-cause mortality (HR, 1.60; 95% CI, 1.04 to 2.46), and therelationship with CVD was essentially unchanged (HR, 1.55; 95%CI, 0.87 to 2.77).
Interactions
The interaction between gender and HbA1c was NS in models thatexamined all-cause (P = 0.35) or CVD mortality (P = 0.15). Therewas no interaction between study as a marker of severity ofkidney disease and HbA1c in models that examined all-cause (P= 0.60) or CVD mortality (P = 0.87).
In this large cohort of nondiabetic patients with CKD, HbA1c,a marker of impaired glucose metabolism, is a significant predictorof all-cause mortality and shows a trend toward being associatedwith CVD mortality. HbA1c concentration reflects average bloodglucose concentration over 3 mo and is a sensitive and reliablemarker of glucose metabolism (14). Cross-sectional studies innondiabetic individuals have shown a relationship between HbA1cand prevalent coronary artery disease as well as markers ofsubclinical atherosclerosis (18,23,24). Population-based prospectivestudies have also demonstrated an association between HbA1cvalues in the nondiabetic range and CVD mortality (3,15,16).Conversely, a nested case-control analysis derived from theWomen’s Health Study cohort found that although HbA1cwas a strong predictor of a composite outcome of fatal and nonfatalCVD events, this relationship was NS after adjustment for otherCVD risk factors (17). No studies, to our knowledge, have examinedthe relationship between HbA1c and outcomes in a CKD population.
Our results suggest that HbA1c is a predictor of all-cause mortalityin nondiabetic CKD and that this relationship is independentof other established CVD risk factors, including CRP, and kidneydisease–related factors. There was a trend toward significancein the relationship between HbA1c and CVD mortality that persistedafter adjustment for potential confounders. The failure to showa relationship between HbA1c and CVD mortality may be due toinadequate statistical power because there were fewer CVD events.
Previous studies using Third National Health and Nutrition ExaminationSurvey data have shown a high prevalence of hyperglycemia inindividuals with GFR < 60 ml/min per 1.73 m2 or microalbuminuria(11,12). Our results therefore suggest that "relative" hyperglycemiamay be an important risk factor contributing to the excess riskof CVD seen in nondiabetic CKD. Furthermore, the persistenceof the association between HbA1c and mortality after exclusionof patients with HbA1c > 6.4% suggests that in CKD, hyperglycemiabelow currently defined thresholds may be associated with adverseoutcomes.
Several studies have suggested a gender difference in the associationbetween HbA1c and mortality. Analyses from the Framingham HeartStudy and the Rancho Bernardo Study found that HbA1c levelsin the nondiabetic range were related to CVD in women but notin men (13,18). In the MDRD Study cohort, there was no interactionbetween gender and all-cause or CVD mortality with HbA1c. Althoughwe cannot be sure, differences between the study populationsin factors such as number of postmenopausal women, CVD riskprofile, and presence of kidney disease itself may account forthe inconsistency of the results. It is also possible that wehad limited power to detect the interaction of HbA1c and gender.
Several pathophysiologic mechanisms may mediate the toxic effectsof chronic hyperglycemia. A growing body of evidence suggestsa link between chronic hyperglycemia and oxidative stress, endothelialdysfunction, and inflammation. HbA1c is a target for intracellularglycoxidation and peroxidation reactions that result in theformation of advanced glycation end products (AGE) (25). TheseAGE have been implicated in the initiation and progression ofatherosclerosis. Chronic hyperglycemia has also been associatedwith increased circulating levels of oxidized LDL, a highlyatherogenic form of LDL cholesterol (26). In a population-basedstudy, moderate elevations in glucose levels were associatedwith abnormalities of cellular antioxidant mechanisms (27).In a study of young healthy Chinese subjects, glucose levelsat the higher end of the normal range were associated with impairedendothelial function and higher carotid intima-media thickness(28). Impaired glucose tolerance has also been associated withan elevated white cell count, which may be a surrogate for chronicinflammation (29). In a cross-sectional study of patients withcoronary atherosclerosis, HbA1c levels in the high normal rangewere associated with higher levels of several inflammatory markers,including CRP, erythrocyte sedimentation rate, and white bloodcell count (30), although in our study, the relationship ofHbA1c with mortality was independent of CRP.
Alternatively, there may be residual confounding from the knownclustering of hyperglycemia with CVD risk factors and othercomponents of the metabolic syndrome, including dyslipidemia,obesity, and hypertension. It is also possible that high baselineHbA1c represents patients who are at higher risk for developingdiabetes in the future and that this accounts for the associationbetween HbA1c and mortality. We do not have data on diabeticstatus during long-term follow-up and therefore are unable toassess this hypothesis. However, this does not preclude theuse of HbA1c as a risk stratification tool for the early identificationof high-risk individuals.
In addition to its relationship with CVD mortality, the associationbetween HbA1c and all-cause mortality may reflect the relationshipbetween abnormalities of glucose metabolism and cancer. Severalpopulation-based studies have shown an association of impairedglucose tolerance and diabetes with mortality from cancers ofthe breast, prostate, colon, pancreas, liver, and gallbladder(31–34). The biologic link between hyperglycemia and thedevelopment of cancer may involve stimulation of IGF-1, whichhas been shown to promote tumor cell growth (35,36).
HbA1c has a few advantages over fasting plasma glucose as amarker of hyperglycemia. First, measurement of HbA1c does notrequire fasting or glucose load. Second, HbA1c provides a morecomprehensive picture of glycemic status and is more indicativeof chronic hyperglycemia than a single plasma glucose measurement.Third, HbA1c reflects both fasting and postprandial hyperglycemia.Several studies have shown that the latter may be a better prognosticmarker than fasting hyperglycemia (37,38). Fourth, HbA1c isnot subject to the wide variations that are inherent in singlemeasures of blood glucose (14).
The strengths of this study include the large number of patientswith nondiabetic kidney disease, long-term follow-up, low prevalenceof CVD at baseline, detailed ascertainment of potentially confoundingvariables, and precise measurement of GFR as a marker of kidneyfunction. The main limitations are lack of follow-up HbA1c measurementsand the potential misclassification of cause of deaths; however,the latter should not influence the relationship between HbA1cand all-cause mortality. In addition, it must be acknowledgedthat the use of the HPLC method to measure HbA1c may resultin overestimation of HbA1c levels in uremia as a result of interferencewith carbamylated hemoglobin (39). However, it has been shownthat carbamylated hemoglobin constitutes a small fraction ofHbA1c in dialysis patients (40).
In summary, HbA1c is an independent predictor of all-cause mortalityin nondiabetic CKD. The association between HbA1c and CVD mortalityneeds further assessment in this population. Larger studieswill allow the evaluation of a threshold in the relationshipbetween level of HbA1c and risk for all-cause and CVD mortality.The results of this study may have important implications: (1)The high prevalence of dysglycemia in CKD may partly explainthe high risk for CVD in CKD, (2) the current definitions fornormal glycemic status may not be appropriate for this population,(3) there may be a role for HbA1c in risk stratification andearly identification of patients who have nondiabetic CKD andare at high risk for CVD, and (4) hyperglycemia may be a potentialtherapeutic target to reduce CVD risk in this patient population.
Acknowledgments
This study was funded by National Institute of Diabetes andDigestive and Kidney Disease grants K23 DK67303, K23 DK02904,and UO1 DK35073.
Footnotes
Published online ahead of print. Publication date availableat www.jasn.org.
Yarnell JW, Pickering JE, Elwood PC, Baker IA, Bainton D, Dawkins C, Phillips DI: Does non-diabetic hyperglycemia predict future IHD? Evidence from the Caerphilly and Speedwell studies.
J Clin Epidemiol 47
: 383
–388, 1994[CrossRef][Medline]
Balkau B, Shipley M, Jarrett R, Pyorala K, Pyorala M, Forhan A, Eschwege E: High blood glucose concentration is a risk factor for mortality in middle-aged nondiabetic men. 20-year follow-up in the Whitehall Study, the Paris Prospective Study, and the Helsinki Policemen Study.
Diabetes Care 21
: 360
–367, 1998[Abstract]
de Vegt F, Dekker JM, Ruhe HG, Stehouwer CD, Nijpels G, Bouter LM, Heine RJ: Hyperglycaemia is associated with all-cause and cardiovascular mortality in the Hoorn population: The Hoorn Study.
Diabetologia 42
: 926
–931, 1999[CrossRef][Medline]
Levitan EB, Song Y, Ford ES, Liu S: Is nondiabetic hyperglycemia a risk factor for cardiovascular disease? A meta-analysis of prospective studies.
Arch Intern Med 164
: 2147
–2155, 2004[Abstract/Free Full Text]
Sasso FC, Carbonara O, Nasti R, Campana B, Marfella R, Torella M, Nappi G, Torella R, Cozzolino D: Glucose metabolism and coronary heart disease in patients with normal glucose tolerance.
JAMA 291
: 1857
–1863, 2004[Abstract/Free Full Text]
K/DOQI clinical practice guidelines for chronic kidney disease: Evaluation, classification, and stratification. Kidney Disease Outcome Quality Initiative.
Am J Kidney Dis 39
: S1
–246, 2002[CrossRef][Medline]
Sarnak MJ, Levey AS, Schoolwerth AC, Coresh J, Culleton B, Hamm LL, McCullough PA, Kasiske BL, Kelepouris E, Klag MJ, Parfrey P, Pfeffer M, Raij L, Spinosa DJ, Wilson PW: Kidney disease as a risk factor for development of cardiovascular disease: A statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention.
Circulation 108
: 2154
–2169, 2003[Free Full Text]
Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY: Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization.
N Engl J Med 351
: 1296
–1305, 2004[Abstract/Free Full Text]
Longenecker JC, Coresh J, Powe NR, Levey AS, Fink NE, Martin A, Klag MJ: Traditional cardiovascular disease risk factors in dialysis patients compared with the general population: The CHOICE Study.
J Am Soc Nephrol 13
: 1918
–1927, 2002[Abstract/Free Full Text]
Sarnak MJ, Coronado BE, Greene T, Wang SR, Kusek JW, Beck GJ, Levey AS: Cardiovascular disease risk factors in chronic renal insufficiency.
Clin Nephrol 57
: 327
–335, 2002[Medline]
Chen J, Muntner P, Hamm LL, Jones DW, Batuman V, Fonseca V, Whelton PK, He J: The metabolic syndrome and chronic kidney disease in US adults.
Ann Intern Med 140
: 167
–174, 2004[Abstract/Free Full Text]
Palaniappan L, Carnethon M, Fortmann SP: Association between microalbuminuria and the metabolic syndrome: NHANES III.
Am J Hypertens 16
: 952
–958, 2003[CrossRef][Medline]
Park S, Barrett-Connor E, Wingard DL, Shan J, Edelstein S: GHb is a better predictor of cardiovascular disease than fasting or postchallenge plasma glucose in women without diabetes. The Rancho Bernardo Study.
Diabetes Care 19
: 450
–456, 1996[Abstract]
Dunn PJ, Cole RA, Soeldner JS, Gleason RE: Reproducibility of hemoglobin AIc and sensitivity to various degrees of glucose intolerance.
Ann Intern Med 91
: 390
–396, 1979
Khaw KT, Wareham N, Luben R, Bingham S, Oakes S, Welch A, Day N: Glycated haemoglobin, diabetes, and mortality in men in Norfolk cohort of European prospective investigation of cancer and nutrition (EPIC-Norfolk).
BMJ 322
: 15
–18, 2001[Abstract/Free Full Text]
Khaw K-T, Wareham N, Bingham S, Luben R, Welch A, Day N: Association of hemoglobin A1c with cardiovascular disease and mortality in adults: The European Prospective Investigation into Cancer in Norfolk.
Ann Intern Med 141
: 413
–420, 2004[Abstract/Free Full Text]
Blake GJ, Pradhan AD, Manson JE, Williams GR, Buring J, Ridker PM, Glynn RJ: Hemoglobin A1c level and future cardiovascular events among women.
Arch Intern Med 164
: 757
–761, 2004[Abstract/Free Full Text]
Singer DE, Nathan DM, Anderson KM, Wilson PW, Evans JC: Association of HbA1c with prevalent cardiovascular disease in the original cohort of the Framingham Heart Study.
Diabetes 41
: 202
–208, 1992[Abstract]
Greene T, Bourgoignie JJ, Habwe V, Kusek JW, Snetselaar LG, Soucie JM, Yamamoto ME: Baseline characteristics in the Modification of Diet in Renal Disease Study.
J Am Soc Nephrol 4
: 1221
–1236, 1993[Abstract]
Standards of medical care in diabetes.
Diabetes Care 27[Suppl 1]
: S15
–S35, 2004
Barrett-Connor EL, Cohn BA, Wingard DL, Edelstein SL: Why is diabetes mellitus a stronger risk factor for fatal ischemic heart disease in women than in men? The Rancho Bernardo Study.
JAMA 265
: 627
–631, 1991[Abstract]
Kuusisto J, Mykkanen L, Pyorala K, Laakso M: NIDDM and its metabolic control predict coronary heart disease in elderly subjects.
Diabetes 43
: 960
–967, 1994[Abstract]
Vitelli LL, Shahar E, Heiss G, McGovern PG, Brancati FL, Eckfeldt JH, Folsom AR: Glycosylated hemoglobin level and carotid intimal-medial thickening in nondiabetic individuals. The Atherosclerosis Risk in Communities Study.
Diabetes Care 20
: 1454
–1458, 1997[Abstract]
Kato T, Chan MC, Gao SZ, Schroeder JS, Yokota M, Murohara T, Iwase M, Noda A, Hunt SA, Valantine HA: Glucose intolerance, as reflected by hemoglobin A1c level, is associated with the incidence and severity of transplant coronary artery disease.
J Am Coll Cardiol 43
: 1034
–1041, 2004[Abstract/Free Full Text]
Himmelfarb J, Stenvinkel P, Ikizler TA, Hakim RM: The elephant in uremia: Oxidant stress as a unifying concept of cardiovascular disease in uremia.
Kidney Int 62
: 1524
–1538, 2002[CrossRef][Medline]
Kopprasch S, Pietzsch J, Kuhlisch E, Fuecker K, Temelkova-Kurktschiev T, Hanefeld M, Kuhne H, Julius U, Graessler J: In vivo evidence for increased oxidation of circulating LDL in impaired glucose tolerance.
Diabetes 51
: 3102
–3106, 2002[Abstract/Free Full Text]
Menon V, Ram M, Dorn J, Armstrong D, Muti P, Freudenheim JL, Browne R, Schunemann H, Trevisan M: Oxidative stress and glucose levels in a population-based sample.
Diabet Med 21
: 1346
–1352, 2004[CrossRef][Medline]
Thomas GN, Chook P, Qiao M, Huang XS, Leong HC, Celermajer DS, Woo KS: Deleterious impact of "high normal" glucose levels and other metabolic syndrome components on arterial endothelial function and intima-media thickness in apparently healthy Chinese subjects: The CATHAY study.
Arterioscler Thromb Vasc Biol 24
: 739
–743, 2004[Abstract/Free Full Text]
Ohshita K, Yamane K, Hanafusa M, Mori H, Mito K, Okubo M, Hara H, Kohno N: Elevated white blood cell count in subjects with impaired glucose tolerance.
Diabetes Care 27
: 491
–496, 2004[Abstract/Free Full Text]
Gustavsson CG, Agardh CD: Markers of inflammation in patients with coronary artery disease are also associated with glycosylated haemoglobin A1c within the normal range.
Eur Heart J 25
: 2120
–2124, 2004[Abstract/Free Full Text]
Gapstur SM, Gann PH, Lowe W, Liu K, Colangelo L, Dyer A: Abnormal glucose metabolism and pancreatic cancer mortality.
JAMA 283
: 2552
–2558, 2000[Abstract/Free Full Text]
Saydah SH, Loria CM, Eberhardt MS, Brancati FL: Abnormal glucose tolerance and the risk of cancer death in the United States.
Am J Epidemiol 157
: 1092
–1100, 2003[Abstract/Free Full Text]
Saydah SH, Platz EA, Rifai N, Pollak MN, Brancati FL, Helzlsouer KJ: Association of markers of insulin and glucose control with subsequent colorectal cancer risk.
Cancer Epidemiol Biomarkers Prev 12
: 412
–418, 2003[Abstract/Free Full Text]
Coughlin SS, Calle EE, Teras LR, Petrelli J, Thun MJ: Diabetes mellitus as a predictor of cancer mortality in a large cohort of US adults.
Am J Epidemiol 159
: 1160
–1167, 2004[Abstract/Free Full Text]
Ma J, Pollak MN, Giovannucci E, Chan JM, Tao Y, Hennekens CH, Stampfer MJ: Prospective study of colorectal cancer risk in men and plasma levels of insulin-like growth factor (IGF)-I and IGF-binding protein-3.
J Natl Cancer Inst 91
: 620
–625, 1999[Abstract/Free Full Text]
Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J, Wilkinson P, Hennekens CH, Pollak M: Plasma insulin-like growth factor-I and prostate cancer risk: A prospective study.
Science 279
: 563
–566, 1998[Abstract/Free Full Text]
Shaw JE, Hodge AM, de Courten M, Chitson P, Zimmet PZ: Isolated post-challenge hyperglycaemia confirmed as a risk factor for mortality.
Diabetologia 42
: 1050
–1054, 1999[CrossRef][Medline]
Hanefeld M, Fischer S, Julius U, Schulze J, Schwanebeck U, Schmechel H, Ziegelasch HJ, Lindner J: Risk factors for myocardial infarction and death in newly detected NIDDM: The Diabetes Intervention Study, 11-year follow-up.
Diabetologia 39
: 1577
–1583, 1996[CrossRef][Medline]
Fluckiger R, Harmon W, Meier W, Loo S, Gabbay KH: Hemoglobin carbamylation in uremia.
N Engl J Med 304
: 823
–827, 1981[Medline]
Engbaek F, Christensen SE, Jespersen B: Enzyme immunoassay of hemoglobin A1c: Analytical characteristics and clinical performance for patients with diabetes mellitus, with and without uremia.
Clin Chem 35
: 93
–97, 1989[Abstract/Free Full Text]
Received for publication May 26, 2005.
Accepted for publication August 30, 2005.
This article has been cited by other articles:
N. Brewer, C. S. Wright, N. Travier, C. W. Cunningham, J. Hornell, N. Pearce, and M. Jeffreys A New Zealand Linkage Study Examining the Associations Between A1C Concentration and Mortality
Diabetes Care,
June 1, 2008;
31(6):
1144 - 1149.
[Abstract][Full Text][PDF]
K. Kalantar-Zadeh, J. D. Kopple, D. L. Regidor, J. Jing, C. S. Shinaberger, J. Aronovitz, C. J. McAllister, D. Whellan, and K. Sharma A1C and Survival in Maintenance Hemodialysis Patients
Diabetes Care,
May 1, 2007;
30(5):
1049 - 1055.
[Abstract][Full Text][PDF]
Top
Abstract
Introduction
2 test, one-way ANOVA, and the Kruskall-Wallis test as appropriate. 
