Download 1001067

Impact of atherosclerosis on the relationship of glycemic control and mortality in diabetic
patients on hemodialysis
Masaaki Inaba1, Kiyoshi Maekawa3, Senji Okuno3, Yasuo Imanishi1, Yasuaki Hayashino4,
Masanori Emoto1, Tetsuo Shoji1, Eiji Ishimura2, Tomoyuki Yamakawa3, Yoshiki Nishizawa1
#1Department of Metabolism, Endocrinology and Molecular Medicine and
#2Department of Nephrology, Osaka City University Graduate School of Medicine, Osaka,
Japan; #3Kidney Center, Shirasagi Hospital, Osaka, Japan; #4 Department of Epidemiology
and Healthcare Research, Kyoto University Graduate School of Medicine and Public Health,
Kyoto.
TEL +81-6-6645-3805, FAX +81-6-6645-3808, e-mail: [email protected]
Runnnig Head: Increased glycoalbumin as risk in DM HD patients
Keywords: glycoalbumin; atherosclerosis; cardiovascular disease; hemodialysis; mortality
Word number of text: 2908
Short summary
The relationship between improved glycemic control and increased survival is dependent on
the presence of cardiovascular disease in DM hemodialysis patients.
1
Abstract
Objective: The impact of pre-existing cardiovascular disease (CVD) on glycemic
control-improved survival in hemodialysis patients with diabetes mellitus (DM) was
investigated. Glycoalbumin (GA) was used as a glycemic marker.
Methods: A single-center 4-year follow-up study was performed in an observational cohort
of 178 DM hemodialysis patients to analyze the relationship between GA and all-cause
mortality in patients with (n=70) and without (n=108) CVD. The subjects were divided into
three categories based on GA value at the start of study.
Results: Baseline characteristics did not differ between two groups of patients. During
4-year follow-up, 24 of 108 (23.3%) CVD(-) patients and 30 of 70 (42.8%) CVD(+) patients
died. The mortality was significantly higher in CVD(+) group. Multivariate Cox analyses
including GA, logCRP, age, gender, hemodialysis duration, albumin, hemoglobin,
BMI, SBP, DBP, smoking habit, and SUN as independent variables showed that GA, in
addition to logCRP and age, was independently associated with mortality in whole
patients. Kaplan-Meier analysis showed lower GA levels as a significant predictor of
lower mortality in CVD(-) group, but not in CVD(+) group. Multivariable-adjusted Cox
proportional hazards models demonstrated a significant association between GA
with all-cause mortality risk in CVD(-) group (p=0.004), in contrast with CVD(+) group
in the same model (p=0.842).
Conclusion: These results demonstrate a beneficial effect of improved glycemic
control on survival in DM hemodialysis patients, which might be attenuated by the
presence of CVD.
2
Introduction
Strict glycemic control decreases diabetes complications, as shown in type 1 diabetes
mellitus (DM) in the Diabetes Control and Complications Trial (1) and in type 2 DM in the
Kumamoto Study (2) and the U.K. Prospective Diabetes Study (3). Strict glycemic control
also has favorable effects on cardiovascular disease (CVD) in DM patients (4-6). It is also
increasingly recognized that the risk of cardiovascular events increases with progression of
the stage of chronic kidney disease (CKD) before initiation of renal replacement therapy
(7,8). We have previously shown that arterial wall thickness (9) and stiffness (10,11), which
are both clinically reliable predictors of mortality in the hemodialysis population (12,13), are
increased in predialysis CKD patients and in patients on maintenance hemodialysis (14,15).
Furthermore, coronary angiography shows a significantly higher prevalence of coronary
artery disease (CAD) in DM patients compared to non-DM patients at the time of initiation of
maintenance hemodialysis (16,17). Multivariate Cox regression analysis has shown that the
presence of CAD is a much stronger risk for cardiovascular death than the presence of DM,
although each factor is significantly associated with development of an initial major adverse
cardiac event in hemodialysis patients (18).
There have been several well-designed studies of the impact of glycemic control on the
prognosis of DM hemodialysis patients, but most have used glycated hemoglobin (HbA1c)
as a marker for glycemic control (19-21). We recently showed that HbA1c can be reduced
independently of glycemic control and is inversely correlated with the weekly dose of
erythropoietin injection (22). Thus we (22,23) and others (24) have proposed glycoalbumin
(GA) as a better index of glycemic control in DM hemodialysis patients. This background
prompted us to determine the impact of CVD at baseline on the effect of glycemic control on
mortality in a cohort of DM hemodialysis patients with and without pre-existing CVD, using
GA as the glycemic index.
Subjects and Methods
3
Study design and subjects
An observational single-center cohort study was performed in DM patients on
maintenance hemodialysis to analyze predictors of mortality. Baseline examinations were
performed in April 2005. Information was collected for pre-existing CVD, medications, body
weight, height, blood pressure, GA, blood chemistry and other clinical parameters. The
cohort was followed until March 2009. At baseline, all patients had been treated by regular
hemodialysis for >1 month at the Kidney Center, Shirasagi Hospital, Osaka, Japan. A total of
178 type 2 DM hemodialysis patients were recruited from the outpatient clinic. The diagnosis
of DM was based on a history of diabetes or on the criteria in the Report of the Expert
Committee on the Diagnosis and Classification of Diabetes Mellitus (25). Due to the small
number of patients with type 1 diabetes in Japan, inclusion of patients with type 1 diabetes
was negated only by a past history of diabetes (26). All patients received 3 to 5 h of
hemodialysis three times a week using standard bicarbonate dialysate. The study was
approved by the Ethics Committee of the Hospital and all subjects gave informed consent to
participation in the study.
Pre-existing CVD at baseline
The presence of CVD at baseline was evaluated using clinical information for coronary,
cerebral and peripheral artery diseases and aortic aneurysm (27). Coronary artery disease
was diagnosed when a subject met one or more of the following criteria: (i) history of
percutaneous coronary intervention or coronary artery bypass grafting, (ii) significant
stenosis on coronary angiography, (iii) ST-T abnormalities on an electrocardiogram
associated with typical symptoms attributable to angina pectoris, and (iv) use of one or more
medications for coronary ischemia. Based on these criteria, 51 patients had coronary artery
disease. Cerebrovascular disease was diagnosed in 21 patients based on past history and
positive findings of infarction or bleeding in X-ray computed tomography (CT) or magnetic
resonance imaging. Peripheral artery disease was diagnosed in 4 patients who had
4
undergone leg amputation due to leg ischemia shown angiographically. No patient was
diagnosed with aortic aneurysm on X-ray CT. At baseline, 70 patients had one or more
cardiovascular complications.
Body weight and blood pressure measurements
Body mass index (BMI) was calculated by dividing dry weight (kg) by squared height (m2).
Blood pressure was measured with a standard mercury sphygmomanometer. Cuffs were
adapted to arm circumference after rest in the supine position for at least 5 min. The systolic
and diastolic blood pressure were taken as the points of appearance and disappearance of
Korotkoff sounds, respectively. The average of three measurements was used for analysis.
Blood sampling and assays
Blood was taken just before the start of the first dialysis session of the week in a supine
position, and drawn into vacuum plastic tubes. Whole blood was used for hemoglobin
determination and serum was used for GA and other biochemical assays. Except for GA,
these assays were performed without delay using an automated analyzer.
Assay for GA
GA was measured as described previously (22) using an enzymatic method in the
Lucica® GA-L kit (Asahi Kasei Pharma Corp., Tokyo, Japan) (28). GA was hydrolyzed to
amino acids by albumin-specific proteinase and then oxidized by ketoamine oxidase to
produce hydrogen peroxide, which was measured quantitatively. The GA level was
calculated as the percentage of glycated albumin relative to total albumin measured in the
same serum sample using the bromocresol purple method (28). The GA assay is unaffected
by physiological concentrations of ascorbic acid, bilirubin, and glucose up to 1,000 mg/dL.
Outcome data collection
5
The hemodialysis cohort was followed until the end of March 2009. At the end of the
follow-up period, 124 patients were alive on hemodialysis and 54 patients had died. The
date and cause of death were obtained by reviewing hospital records. Deaths from cardiac,
cerebrovascular, and peripheral vascular diseases were categorized as deaths due to
cardiovascular disease and sepsis; and those from pneumonia, enteritis, and other diseases
of bacteria or fungi were categorized as deaths due to infectious disease. Five patients with
CVD had been lost from the study during the study period due to the movement to another
Hospital.
Statistical analysis
Continuous variables are shown as means ± SD. Differences in means and
percentages between two groups were evaluated by analysis of variance and χ2 test,
respectively. Survival curves were constructed by the Kaplan-Meier method and
differences between groups evaluated by log-rank test. Prognostic variables for
survival were examined using univariate Cox proportional hazards regression models
and independent predictors of death were determined using multivariate Cox
analyses. Next, to evaluate the hypothesis that the effect of GA on mortality may vary
depending on baseline comorbidity of CVD, we evaluated the joint association of
presence of absence of CVD and GA categories with all-cause mortality. Likelihood
ratio tests were used to examine statistical interactions between CVD and GA
categories by comparing the −2 log(likelihood) between two nested models, one with
only the main effects, and the other with the main effects and interaction terms
between CVD and GA categories. We used two statistical models for this analysis;
first model adjusted for age, gender hemodialysis duration, serum albumin,
hemoglobin, BMI, SBP, DBP, SUN; second model adjusted for all variables in the first
model and CVD comorbidities. Proportional hazard assumptions were confirmed with
visual inspection, in which we confirmed that the graph of the survival function
6
versus the survival time resulted in a graph with parallel curves, and similarly that the
graph of the log(−log(survival)) versus log of survival time graph resulted in parallel
lines [14]. P < 0.05 was considered significant. All analyses were performed using
Stata/SE 11.1 (Stata Corporation, College Station, TX, USA).
Results
Baseline characteristics
The characteristics of the cohort in May 2005 are summarized in Table 1. At baseline, 70
of the 178 DM hemodialysis patients had CVD. There were no significant differences
between patients with and without pre-existing CVD in terms of gender, age, hemodialysis
duration, BMI, blood pressure, albumin, hemoglobin, and SUN, which are all established risk
factors for mortality in hemodialysis patients. Similarly, GA did not differ significantly between
patients with (23.3±6.2%) and without (23.5±6.2%) pre-existing CVD.
Outcomes and mortality
In the total cohort, 54 deaths occurred during the follow-up period of 48 months. The
crude annual mortalities were 10.7% and 5.8% in patients with and without pre-existing
CVD, respectively. The mortality was significantly higher in the CVD(+) group than in the
CVD(-) group. Also, mortality was higher in the CVD(+) group and lower in the CVD(-) group
compared to the mortality of 9.4% for all Japanese dialysis patients in 1996 reported by the
Japanese Society for Dialysis Therapy. The 24 deaths in CVD(-) patients included 13 from
cardiovascular disease [coronary heart disease (n=4), cerebrovascular disease (n=5), and
sudden death (n=4)] and 11 from non-cardiovascular causes [infectious disease (n=3),
malignancy (n=4), respiratory failure (n=1), generalized weakness (n=2), and suicide (n=1)].
The 30 deaths in CVD(+) patients were due to coronary heart disease (n=2),
cerebrovascular disease (n=6), congestive heart failure (n=8), ischemic colitis (n=1), sudden
death (n=5), infectious disease (n=6), malignancy (n=1), and generalized weakness (n=1).
7
Mortality rates in whole DM hemodialysis patients
To elucidate the association of GA with mortality in DM hemodialysis patients,
multivariable-adjusted Cox analysis of the entire cohort as a single population was
performed (Table 2). GA showed an independent and significant association with
mortality after adjustment for gender, hemodialysis duration, serum albumin,
hemoglobin, BMI, SBP, DBP, smoking habit, and BUN, even with or without inclusion
of CVD as an independent variable.
Since the inclusion of CVD as an independent
variable as shown in Model 2 in Table 2 did not affect at all hazard ratio of 1.042 of
Model 1 which did not include CVD as an independent variable. The magnitude of
association between GA and mortality was not attenuated by adding CVD into the
statistical model, which clearly demonstrated that CVD prevalence is not an
intermediate variable to affect the association of GA and mortality in hemodialysis
patients.
Mortality rates in DM hemodialysis patients with and without pre-existing CVD
We next analyzed the association between GA and mortality in either CVD(-) or
CVD(+) hemodialysis patients, by the Kaplan–Meier method (Figure 1). Each group of
patient was placed into three categories on the basis of GA values. The tertiles are referred
to as T1 (lowest GA), T2, and T3, respectively. The ranges of GA among these categories
were <20.0% (T1), 20.0-24.5% (T2), and ≥24.5% (T3). The numbers of patients in the three
categories was 34, 36, and 38 for the CVD(-) group and 25, 23, and 22 for the CVD(+) group.
For CVD(-) patients, mortality was significantly lower in the T1 category than in T2 or T3
(P=0.026), whereas mortality did not differ significantly among the three categories for
CVD(+) patients.
8
Multivariate Cox proportional hazards analysis in CVD(−) and CVD(+) patients
We examined whether the significant association between GA and mortality in the
CVD(-) group was independent of other confounding variables using multivariate Cox
models (Table 3). After adjustment for known risk factors for mortality in hemodialysis
patients (age, Log CRP, gender, hemodialysis duration, serum albumin, hemoglobin,
BMI, SBP, DBP, smoking habit, and SUN), a higher GA level was still significantly
associated with a higher mortality in CVD(-) patients [Hazard ratio: 1.042(95% CI:
1.029-1.160, p=0.004)]. However, GA did not show a significant association with
mortality in CVD(+) patients [Hazard ratio: 1.006 (95% CI: 0.945-1.071, p=0.842)],
although the significant association mortality with log CRP was retained. The
difference in the association between CVD(-) and CVD(+) patients showed a borderline
significance (p for interaction=0.0608), although not significantly different, suggesting
that pre-existing CVD is an effect modifier of the association between GA and mortality
in DM hemodialysis patients.
.
Discussion
In this study, we examined whether pre-existing CVD affects the impact of glycemic
control on mortality in DM hemodialysis patients, using GA as an index of glycemic control. A
significant association between lower GA and reduced mortality was found in patients
without CVD, whereas those with CVD did not show such an association. The different
GA-mortality relationship in the two groups remained evident after multivariate adjustment.
These data indicate that an improvement of glycemic control, as reflected by a lower GA
value, is significantly associated with a lower risk of mortality in DM hemodialysis patients
without CVD, but not in those with CVD. This suggests that improved glycemic control had a
protective effect against death during maintenance hemodialysis in the absence of
pre-existing CVD. Thus, the presence of pre-existing CVD is an important factor that
9
attenuates the beneficial effect of glycemic control on mortality in DM hemodialysis patients.
We (22,23) and others (24) have demonstrated that GA is a more relevant parameter
than HbA1c for assessment of glycemic control in DM hemodialysis patients, because of the
apparent reduction of HbA1c by erythropoiesis-stimulating agents (ESAs) due to the
increased proportion of younger erythrocytes after treatment (25). Measurement of HbA1c in
DM hemodialysis patients leads to a significant underestimation of glycemic control by 33%
on average (22). Furthermore, we have shown that GA, but not HbA1c, is associated with
pulse-wave velocity (30), prevalence of peripheral vascular calcification (31), and the
osteosonographic calcaneal index (32) in DM hemodialysis patients. These data suggest
that improved glycemic control protects against development of DM complications during
hemodialysis. The causality of this relationship cannot be assessed in a cross-sectional
study.
Evidence is accumulating that indicates a close association between poorer glycemic
control and a poor outcome in DM hemodialysis patients (19-21,33), although most previous
studies have been small and have used HbA1c as an index of glycemic control. It is of note
that a recent study of 23,618 DM hemodialysis patients found a paradoxically lower
unadjusted mortality associated with higher HbA1c levels (34). However, after adjusting for
markers of malnutrition and inflammation, higher HbA1c was associated with greater
mortality. Thus, competing risk factors related to malnutrition, wasting, and anemia in these
patients may have confounded the association between glycemic control and survival (34).
Our results suggest that the presence of pre-existing CVD is another important factor that
affects the impact of glycemic control on mortality in DM hemodialysis patients.
Current therapeutic targets of glycemic control in CKD are based on trials performed in
DM patients with normal kidney function.(35) Considering the tradeoff between strict
glycemic control and the greater incidence of hypoglycemia (which occurs more easily in
hemodialysis patients), and the time required to obtain a beneficial effect of better glycemic
control, the current American Diabetes Association clinical practice recommendations allow
10
for less strict glycemic control in patients with shorter life expectancy, possibly including
those on hemodialysis (36). Our results show that a therapeutic target of glycemic control
with GA <20.0% may be appropriate to reduce mortality significantly in DM hemodialysis
patients without CVD (Figure 1). Since a 3% increase in GA is equivalent to a 1% increase in
HbA1c (37), GA of 20.0% is similar to HbA1c of 6.6% in DM patients with normal renal
function, although the measured HbA1c in patients taking ESAs will be lower than 6.6%.
It is of interest that GA and age showed no significant association with mortality in DM
hemodialysis patients with CVD in Cox regression analysis, while log CRP remained
significant. Inflammation as reflected by increased CRP contributes to both plaque formation
and plaque instability (38) and is related to ruptured plaque (39). The effects of aging and
glycemic control may be attenuated more in vessels with greater atherosclerotic changes,
and this may explain why log CRP retained a significant association with mortality in CVD(+)
patients.
There are several limitations in this study. First, because of the observational nature of
the study, associations do not necessarily indicate causality. Therefore, although the results
are of clinical interest, randomized controlled trials are needed to prove the hypothesized
relationships. Second, we cannot define the exact target range for GA required to reduce
mortality in patients without CVD since the results are based on a small scale, single center
study. Third, it is unclear from this study whether much poorer glycemic control would have a
deleterious effect on survival in DM hemodialysis patients with pre-existing CVD. Such an
analysis was difficult to perform because the mean GA values in the cohort were around
23%, indicating good glycemic control compared to previous studies using HbA1c as a
marker. Fourth, although the cohort included all DM hemodialysis patients from a single
hemodialysis center, there might have been some bias since this was a single-center study
of prevalent dialysis patients in an urban area of Osaka, Japan. Thus, generalization of the
results should be considered carefully with regard to ethnic groups, geographic areas and
prevalent versus incident dialysis patients. Fifth, we calculated the mortality risk based on
11
single-point measurements of GA at the start of the study, rather than on averaged values
obtained during follow-up. Therefore, we were only able to show a remote effect on mortality
during 4-year follow-up. Analysis based on sequential measurements during follow-up may
reveal a more intimate association between GA and mortality. Sixth, the definition of
pre-exiting CVD might not reflect the degree of atherosclerosis precisely in
hemodialysis patients. Since many DM hemodialysis patients have asymptomatic
coronary artery disease (16-18) and such patients with advanced atherosclerosis may have
been excluded from the CVD(+) group, the level of glycemic control and mortality risk might
have been underestimated.
In conclusion, the present observational study indicates that improved glycemic control
as reflected by lower GA is only associated with a better prognosis in DM hemodialysis
patients without pre-existing CVD, and not in those with CVD. This implies that optimal
target levels for glycemic control might differ depending on cardiovascular comorbidity.
12
References
1. The Diabetes Control and Complications Trial Research Group: The effect of intensive
treatment of diabetes on the development and progression of long-term complications in
insulin-dependent diabetes mellitus: the Diabetes Control and Complications Trial
Research Group. N Engl J Med 329: 977–986, 1993
2. Ohkubo Y, Kishikawa H, Araki E, Miyata T, Isami S, Motoyoshi S, Kojima Y, Furuyoshi
N, Shichiri M: Intensive insulin therapy prevents the progression of diabetic
microvascular complications in Japanese patients with non-insulin-dependent diabetes
mellitus: a randomized prospective 6-year study. Diabetes Res Clin Pract 28:103–117,
1995
3. UK Prospective Diabetes Study (UKPDS) Group: Intensive blood-glucose control with
sulphonylureas or insulin compared with conventional treatment and risk of
complications in patients with type 2 diabetes (UKPDS 33): UK Prospective Diabetes
Study (UKPDS) Group. Lancet 352:837–853, 1998 [erratum in Lancet354:602, 1999]
4. Gaede P, Vedel P, Larsen N, Jensen GV, Parving HH, Pedersen O: Multifactorial
intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med
348:383–393, 2003
5. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M: Acarbose
treatment and the risk of cardiovascular disease and hypertension in patients with
impaired glucose tolerance: the STOP-NIDDM trial. JAMA 290:486–494, 2003
6. Nathan DM, Lachin J, Cleary P, Orchard T, Brillon DJ, Backlund JY, O’Leary DH,
Genuth S: Intensive diabetes therapy and carotid intima-media thickness in type 1
diabetes mellitus. N Engl J Med 348:2294–2303, 2003
7. Brugts JJ, Knetsch AM, Mattace-Raso FU, et al. Renal function and risk of myocardial
infarction in an elderly population: the Rotterdam Study. Arch Intern Med
165:2659–2665, 2005.
8. Shlipak MG, Sarnak MJ, Katz R, et al. Cystatin C and the risk of death and
13
cardiovascular events among elderly persons. N Engl J Med (2005) 352:2049–2060.
9. Shoji T, Emoto M, Tabata T et al. Advanced atherosclerosis in predialysis patients with
chronic renal failure. Kidney Int 61:2187–2192, 2002.
10. Shinohara K, Shoji T, Tsujimoto Y et al. Arterial stiffness in predialysis patients with
uremia. Kidney Int 65: 936–943, 2004.
11. Kimoto E, Shoji T, Shinohara K et al. Regional arterial stiffness in patients with type 2
diabetes and chronic kidney disease. J Am Soc Nephrol 17: 2245–2252, 2006.
12. Shoji T, Emoto M, Shinohara K et al. Diabetes mellitus, aortic stiffness, and
cardiovascular mortality in end-stage renal disease. J Am Soc Nephrol 12: 2117–2124,
2001.
13. Nishizawa Y, Shoji T, Maekawa K et al. Intima-media thickness of carotid artery predicts
cardiovascular mortality in hemodialysis patients. Am J Kidney Dis 41: S76–S79, 2003.
14. Kawagishi T, Nishizawa Y, Konishi T et al. High-resolution B-mode ultrasonography in
evaluation of atherosclerosis in uremia. Kidney Int 48: 820–826, 1995.
15. London GM, Marchais SJ, Safar ME et al. Aortic and large artery compliance in
end-stage renal failure. Kidney Int 37: 137–142, 1990.
16. Joki N, Hase H, Nakamura R, Yamaguchi T. Onset of coronary artery disease prior to
initiation of haemodialysis in patients with end-stage renal disease. Nephrol Dial
Transplant. 12(4):718-23, 1997.
17. Ohtake T, Kobayashi S, Moriya H, Negishi K, Okamoto K, Maesato K, Saito S. High
prevalence of occult coronary artery stenosis in patients with chronic kidney disease at
the initiation of renal replacement therapy: an angiographic examination. J Am Soc
Nephrol. 16(4):1141-8, 2005.
18. Hase H, Tsunoda T, Tanaka Y, Takahashi Y, Imamura Y, Ishikawa H, Inishi Y, Joki N. Risk
factors for de novo acute cardiac events in patients initiating hemodialysis with no
previous cardiac symptom. Kidney Int. 70(6):1142-8, 2006.
19. Morioka T, Emoto M, Tabata T, Shoji T, Tahara H, Kishimoto H, Ishimura E, Nishizawa Y.
14
Glycemic control is a predictor of survival for diabetic patients on hemodialysis. Diabetes
Care. 24(5):909-13, 2001.
20. Medina RA, Pugh JA, Monterrosa A, Cornell J: Minority advantage in diabetic end-stage
renal disease survival on hemodialysis: due to different proportions of diabetic type? Am
J Kidney Dis 28:226–234, 1996.
21. Hayashino Y, Fukuhara S, Akiba T, Akizawa T, Asano Y, Saito A, Bragg-Gresham JL,
Ramirez SP, Port FK, Kurokawa K. Diabetes, glycaemic control and mortality risk in
patients on haemodialysis: the Japan Dialysis Outcomes and Practice Pattern Study.
Diabetologia. 50(6):1170-7, 2007.
22. Inaba M, Okuno S, Kumeda Y, et al. Glycated albumin is a better glycemic indicator than
glycated hemoglobin values in hemodialysis patients with diabetes: effect of anemia and
erythropoietin injection. J Am Soc Nephrol 18:896-903, 2007.
23. Nagayama H, Inaba M, Okabe R, Emoto M, Ishimura E, Okazaki S, Nishizawa Y.
Glycated albumin as an improved indicator of glycemic control in hemodialysis patients
with type 2 diabetes based on fasting plasma glucose and oral glucose tolerance test.
Biomed Pharmacother. 63(3):236-40, 2009.
24. Peacock TP, Shihabi ZK, Bleyer AJ, Dolbare EL, Byers JR, Knovich MA,
Calles-Escandon J, Russell GB, Freedman BI. Comparison of glycated albumin and
hemoglobin A(1c) levels in diabetic subjects on hemodialysis. Kidney Int. 73(9):1062-8,
2008.
25. American Diabetes Association.: Mellitus TECotDaCoD. Report on the expert committee
on the diagnosis and classification of diabetes mellitus. Diab Care 21(suppl 1), S5-S22,
2005.
26. Inaba M, Nagasue K, Okuno S, Ueda M, Kumeda Y, Imanishi Y, Shoji T, Ishimura E,
Ohta T, Nakatani T, Kim M, Nishizawa Y. Impaired secretion of parathyroid hormone, but
not refractoriness of osteoblast, is a major mechanism of low bone turnover in
hemodialyzed patients with diabetes mellitus. Am J Kidney Dis. 39, 1261-1269, 2002.
15
27. Maekawa K, Shoji T, Emoto M, Okuno S, Yamakawa T, Ishimura E, Inaba M, Nishizawa
Y. Influence of atherosclerosis on the relationship between anaemia and mortality risk in
haemodialysis patients. Nephrol Dial Transplant. 23(7):2329-36, 2008.
28. Kouzuma T Study of glycated amino acid elimination for an improved enzymatic
glycated albumin measurement method. Clin. Chim. Acta 346, 135-143, 2004.
29. Nakao T, Matsumoto H, Okada T, Han M, Hidaka H, Yoshino M, Shino T, Yamada C,
Nagaoka Y. Influence of erythropoietin treatment on hemoglobin A1c levels in patients
with chronic renal failure on hemodialysis. Intern Med 38:826-830, 1998.
30. Kumeda Y, Inaba M, Shoji S, Ishimura E, Inariba H, Yabe S, Okamura M, Nishizawa Y.
Significant correlation of glycated albumin, but not glycated haemoglobin, with arterial
stiffening in haemodialysis patients with type 2 diabetes. Clin Endocrinol (Oxf).
69(4):556-61, 2008.
31. Yamada S, Inaba M, Shidara K, Okada S, Emoto M, Ishimura E, Nishizawa Y.
Association of glycated albumin, but not glycated hemoglobin, with peripheral vascular
calcification in hemodialysis patients with type 2 diabetes. Life Sci. 83(13-14):516-9,
2008.
32. Yamada S, Inaba M, Okada S, Imanishi Y, Mori K, Emoto M, Ishimura E, Nishizawa Y.
Association of glycated albumin, but not glycated hemoglobin, with calcaneus
quantitative ultrasound in male hemodialysis patients with type 2 diabetes mellitus.
Metabolism. in press.
33. Ishimura E, Okuno S, Kono K, Fujino-Kato Y, Maeno Y, Kagitani S, Tsuboniwa N,
Nagasue K, Maekawa K, Yamakawa T, Inaba M, Nishizawa Y. Glycemic control and
survival of diabetic hemodialysis patients--importance of lower hemoglobin A1C levels.
Diabetes Res Clin Pract. 83(3):320-6, 2009.
34. Kalantar-Zadeh K, Kopple JD, Regidor DL, et al: A1C and survival in maintenance
hemodialysis patients. Diabetes Care 30:1049-1055, 2007
35. Writing Team for the Diabetes Control and Complications Trial/Epidemiology of Diabetes
16
Interventions and Complications Research Group: Effect of intensive therapy on the
microvascular complications of type 1 diabetes mellitus. JAMA 287:2563-2569, 2002.
36. American Diabetes Association: Standards of Medical Care in Diabetes—2008.
Diabetes Care 31:S12-S54, 2008 (suppl 1).
37. Tahara Y: Glycoalbumin [in Japanese], Kettouti wo Miru, Kangaeru, 2000, pp 62–69
38. Kashiwagi M, Tanaka A, Kitabata H, Tsujioka H, Matsumoto H, Arita Y, Ookochi K, Kuroi
A, Kataiwa H, Tanimoto T, Ikejima H, Takarada S, Kubo T, Hirata K, Nakamura N,
Mizukoshi M, Imanishi T, Akasaka T. Relationship between coronary arterial remodeling,
fibrous cap thickness and high-sensitivity C-reactive protein levels in patients with acute
coronary syndrome. Circ J. 73(7):1291-5, 2009.
39. Sano T, Tanaka A, Namba M, Nishibori Y, Nishida Y, Kawarabayashi T, Fukuda D,
Shimada K, Yoshikawa J. C-reactive protein and lesion morphology in patients with
acute myocardial infarction. Circulation. 108(3):282-5, 2003.
17
Table 1. Baseline characteristics of DM hemodialysis patients with and without pre-existing
CVD at the start of the study
CVD(-)
CVD(+)
P value
N
108
Gender (M/F)
72/36
Age (year)
65.4±10.6
HD duration (months)
60 ± 54
BMI (kg/m2)
22.7 ± 3.4
Systolic BP (mmHg)
161 ± 24
Diastolic BP (mmHg)
71 ± 24
SUN (mg/dl)
67 ± 14
Serum Alb (g/dl)
3.6 ± 0.4
Hb (g/dl)
10.3 ± 1.1
CRP (mg/dl)
0.37 ± 0.88
GA (%)
23.5 ± 6.2
Data are expressed as the mean  SD.
P < 0.05 was considered to be statistically significant
18
70
-
55/15
66.3 ± 8.1
63 ± 48
23.3 ± 6.2
167 ± 26
67 ± 27
66 ± 16
3.6 ± 0.4
10.4 ± 1.1
0.31± 0.47
23.3± 6.2
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
ns
Table 2 Multivariable-adjusted Cox analysis to examine the association of GA and mortality
in total patients without (Model 1) and with (Model 2) pre-existing CVD as an independent
variable
Model 1*
Model 2 †
Predictors
HR (95% CI)
P-value
HR (95% CI)
P-value
GA (per 1%)
1.042 (1.00-1.086)
0.053
1.045 (1.003-1.090)
0.035
Log CRP
2.084 (1.255-3.462)
0.005
2.501 (1.483-4.218)
0.001
Age (per 1year)
1.037 (1.000-1.087)
0.059
1.030 (0.991-1.071)
0.137
* Adjusted for gender, HD duration, serum albumin, Hb, BMI, SBP, DBP, smoking, SUN
† Adjusted for all variables in model 1 and history of CVD
19
Table 3 Association between GA and mortality stratified by the presence/absence of
pre-existing CVD
CVD(-)
CVD(+)
P-value for
interaction
Predictors
HR (95% CI)
P-value
HR (95% CI)
P-value
1093
1.006
GA (per 1%)
0.004
0.842
0.0608
(1.029-1.160)
(0.945-1.071)
2.242
2.806
Log CRP
0.014
0.005
0.6424
(1.118-4.496)
(1.375-5.726)
1.048
1.006
Age (per 1year)
0.063
0.817
0.2416
(0.997-1.101)
(0.953-1.063)
* Adjusted for gender, HD duration, serum albumin, Hb, BMI, SBP, DBP, smoking, SUN
20
Figure Legends
Figure 1. Kaplan-Meier curves showing the association between GA and all-cause mortality
in hemodialysis patients with and without pre-existing CVD at baseline. P-values were
calculated by log-rank test.
21