Sei sulla pagina 1di 6

Atherosclerosis 225 (2012) 450e455

Contents lists available at SciVerse ScienceDirect

Atherosclerosis
journal homepage: www.elsevier.com/locate/atherosclerosis

Serum glycated albumin predicts the progression of carotid arterial


atherosclerosis
Sun Ok Song a, Kwang Joon Kim a, b, Byung-Wan Lee a, *, Eun Seok Kang a, Bong Soo Cha a, Hyun Chul Lee a
a
Division of Endocrinology and Metabolism, Department of Internal Medicine, Severance Hospital, Yonsei University College of Medicine, 134 Shinchon-dong, Seodaemun-gu,
Seoul 120-752, Republic of Korea
b
Severance Executive Healthcare Clinic, Severance Hospital, Yonsei University Health System, Seoul, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: We investigated the association between the ratio of serum glycated albumin (GA) levels to
Received 22 June 2012 hyperglycemic levels (GA/A1c ratio) and the progression of carotid artery atherosclerosis.
Received in revised form Methods: For this retrospective longitudinal study we recruited patients who had undergone carotid IMT
24 August 2012
measurement twice and been tested consecutively for both A1c and GA levels every 3 or 6 months. The
Accepted 6 September 2012
Available online 18 September 2012
subjects were classified into two groups based on non-progression (group I) and progression (group II) of
carotid IMT. Mean values of A1c and GA and the GA/A1c ratio were compared between groups.
Results: Of the 218 subjects (122 men and 96 women), group II (n¼77) showed significantly higher
Keywords:
Glycated albumin
baseline systolic BP, eGFR, GA, GA/A1c ratio, and progression of carotid IMT (D IMT) than group I (n ¼
Type 2 diabetes 141). The mean A1c level tended to be higher in group II than in group I (p ¼ 0.054). By Spearman's
Carotid intima media correlation test, baseline diastolic BP, total cholesterol, triglyceride, GA, and GA/A1c were significantly
Thickness associated with D IMT in group II but not in group I. In multivariate regression analysis, serum level of GA
GA/A1c ratio and GA/HbA1c ratio predicted progression of IMT after adjustment for other risk factors in both models
applied.
Conclusion: We suggest that glycated albumin is not only a useful glycemic index but also might be an
atherogenic protein in the development of diabetic atherosclerosis.
Ó 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction peripheral arterial disease documented by cardiac MRI and ABI [7].
Although an association between carotid IMT progression and
Atherosclerosis is a major vascular complication in type 2 dia- several common risk variables such as hyperlipidemia, hyperten-
betes (T2D) and can have severe consequences such as coronary sion, and hyperglycemia has been widely demonstrated [8], few
heart disease or stroke. Among the known risk factors for athero- clinical studies have been performed on the role of GA in predicting
sclerosis, there is controversy over the role of advanced glycated diabetic vascular complications.
end product (AGE)-mediated vascular inflammation in the devel- In this study, we hypothesized that, in addition to higher glyce-
opment or progression of atherosclerosis in patients with T2D [1,2]. mic indices, increased glycation indices might also predict or reflect
The utility of glycated albumin (GA) as an intermediate term (2e3 an increase or progression of carotid IMT. Based on this hypothesis,
weeks) parameter for glucose monitoring and as a useful and rapid we aimed to investigate the association between carotid artery
glycation index for glucose fluctuation and excursion is well atherosclerosis and an elevated ratio of GA levels relative to hyper-
recognized [3], but there is growing evidence that GA, a precursor glycemic levels (measured by glycated hemoglobin or A1c levels) by
of AGEs, is also associated with enhanced oxidative stress and adopting the GA/A1c ratio used for controlling hyperglycemia.
endothelial injury resulting in atherosclerosis in diabetic subjects
[2,4,5,6]. In a recent cross-sectional study, we demonstrated that an
increased GA level is associated with atherosclerotic plaque 2. Methods
formation assessed by carotid intima media thickness (IMT) in type
2 diabetic subjects without evidence of coronary heart disease and 2.1. Subjects and research design

This study was a retrospective longitudinal study on type 2 dia-


* Corresponding author. Tel.: þ82 2 2228 1938; fax: þ82 2 393 6884. betic subjects who were registered on the Severance Hospital Dia-
E-mail address: bwanlee@yuhs.ac (B.-W. Lee). betes Complications Registry between March 2009 and April 2012.

0021-9150/$ e see front matter Ó 2012 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.atherosclerosis.2012.09.005
S.O. Song et al. / Atherosclerosis 225 (2012) 450e455 451

We recruited adult subjects who satisfied the following criteria: (1) variables were expressed as medians (interquartile range). Discrete
had undergone carotid IMT measurement twice in this period; and variables were expressed as percentages. Statistical comparisons
(2) had been tested consecutively for both A1c and GA every 3 or 6 between groups with and without carotid artery IMT progression
months. Exclusion criteria included patients unavailable for repeat were performed using the ManneWhitney U test or c2 test which are
GA analysis within a 6-month period. We defined the positive values non-parametric statistical methods. Spearman’s correlation coeffi-
of D IMT (D IMT ¼ follow-up IMT  baseline IMT) as progression of cient, a type of non-parametric correlation analysis, was used to
carotid IMT, and even or negative D IMT as non-progression. Based on determine the relationship between IMT changes and the continuous
this definition, we classified the subjects into non-progression and variables. Multivariate logistic regression analysis was used to esti-
progression of carotid IMT during the time interval of carotid IMT mate multiple correlations between carotid artery IMT changes and
measurements. The study protocol was approved by the ethics clinical and laboratory risk factors. c2 test was used for comparison of
committee of Yonsei University College of Medicine. carotid IMT changes according to the progression or regression of IMT
level. Data with p value <.05 were considered significant.
2.2. Laboratory assessments
3. Results
Blood samples were collected from subjects after an overnight
fast. Plasma glucose was measured using the glucose oxidase 3.1. Characteristics of the study participants
method. Plasma triglycerides, total cholesterol, high-density lipo-
protein cholesterol (HDL cholesterol), blood urea nitrogen, creati- A total of 218 subjects (122 men and 96 women; median age,
nine, AST, and ALT levels were assayed using a routine Hitachi 7600 62.0 [interquartile range (IQR), 53.0e68.0] years) were finally
autoanalyzer (Hitachi Instruments Service, Tokyo, Japan). Low- enrolled in this study. The enrolled subjects were classified into
density lipoprotein cholesterol (LDL cholesterol) was calculated non-progression (group I, n ¼ 141) and progression (group II,
using the Friedewald equation. Serum glycated albumin (GA) was n ¼ 77) of carotid IMT during the time interval of carotid IMT
determined by an enzymatic method using an albumin-specific measurements. The median duration of diabetes was 6.00 (3.00e
proteinase, ketamine oxidase, albumin assay reagents (LUCICA 12.0) years. The median levels of serum A1c, GA, and the value of
GA-L, Asahi Kasei Pharma Co., Tokyo, Japan), and a Hitachi 7699 P GA/A1c were 7.00% (6.40e7.95), 14.6% (16.9e20.9), and 2.17 (2.45e
module autoanalyzer. The coefficient of variation (CV) was 1.43%. 2.76), respectively. The median BMI of the study subjects was 25.3
Serum glycated hemoglobin (HbA1c; henceforth described as A1c) (23.3e27.5) kg/m2. Table 1 shows the demographic and laboratory
was measured by high-performance liquid chromatography (HPLC) characteristics between the two groups, focusing on glucose
using a Variant II Turbo system (Bio-Rad Laboratories, Hercules, metabolism and cardiovascular risk variables. The median age (62
CA). The reference interval was between 4.0 and 6.0% for A1c and (IQR 53.0e68.0) vs. 62 (IQR 53.8e69.3)) and gender (52.5% vs.
between 11.0 and 16.0% for GA. 62.3%) were relatively similar between patients with non-
The average of A1c (henceforth referred to as mean A1c), GA progression (group I) and progression (group II) of IMT, respec-
(mean GA), and GA/A1c ratio (mean GA/A1c) were calculated by the tively. Except for SBP and parameters for glycemic indices, there
sum of every measured values in each indices divided by the were no differences in other anthropometric characteristics
numbers of values throughout the study period. between two groups. Systolic blood pressure was higher in group II
(129 (IQR 119.0e142.3)) than in group I (125 (IQR 118.0e137.0),
2.3. Measurement of intima media thickness (IMT) p ¼ 0.032). Compared to group I without-progression of carotid
IMT, the baseline GA (16.3% (IQR 14.2e20.5) vs. 18.1% (IQR 15.7e
The change in IMT (mm) (D IMT ¼ follow-up IMT  baseline IMT) 23.4), p ¼ 0.016), baseline GA/A1c ratio (2.40 (IQR 2.14e2.69) vs.
during the study period was calculated. IMT measurement has been 2.49 (IQR 30e2.91), p ¼ 0.013), the mean GA over the study period
described in detail in previous reports [7]. Briefly, common carotid (15.8%(IQR 14.1e18.6) vs. 17.5% (IQR 14.8e21.7), p ¼ 0.011) and
arterial ultrasound examinations were conducted by two specialized mean GA/A1c ratio (2.34 (IQR 2.13e2.54) vs. 2.49 (IQR 2.19e2.85),
technicians using an Aloka ProSound ALPHA 10 (HITACHI, Tokyo, p ¼ 0.019) were significantly higher in group II with progression of
Japan) with a 13 MHz linear probe. All measurements were recorded carotid IMT. However, baseline A1c (6.8% (IQR 6.4e7.8), 7.2% (IQR
with the subject in a supine position with the head elevated to 45 6.6e8.1), p ¼ 0.157) and mean A1c across the study period (6.9%
and tilted to either side by 30 . We performed B mode examinations (IQR 6.5e7.5) vs. 7.08 (IQR 7.7e7.9), p ¼ 0.054) were not signifi-
1.5 cm proximal to the carotid bifurcation on the far wall of the cantly different between the groups (Table 1 and Fig. 2). Regarding
common carotid artery and on both sides. IMT was defined as the to the correlation of GA with A1c, which is the golden standard for
distance between the mediaeadventitia interface and the lumene glycemic index, a strong linear relationship was observed between
intima interface. Average IMT was the mean value of computer-based mean serum GA and mean A1c levels in all patients in the
points in the region, and maximal IMT was the IMT value at the progression and non-progression groups (r ¼ .767, .749 and .783,
maximal point of the region [9]. Reproducibility was further tested respectively, p < 0.001) (Fig. 1).
for carotid IMT measurements: inter-observer coefficient of variation We also ascertained the subjects rank in each percentile for
(CV) was 2.11%, and the day-to-day coefficient of variation was 2.06%. baseline A1c (10, 25, 50, 75, 90) and compared glycemic indices,
Plaques were defined according to the Mannheim consensus [10], in baseline GA, baseline GA/A1c, mean GA, and mean GA/A1c between
which a plaque is diagnosed when the vessel wall thickness is greater group I and group II (Supplementary Figure S1). Baseline GA, mean
than 1.5 mm or when the vessel wall appears to be at least .5 mm or GA/A1c, mean GA, and mean GA/A1c were higher in group II than in
50% thicker than the surrounding wall. group I for every percentile of baseline HbA1c.

2.4. Statistical analysis 4. Association of carotid IMT progression with glycemic


indices and cardiovascular risk variables
All statistical analyses were performed with PASW statistics soft-
ware (version 18.0; SPSS Inc., Chicago, IL). We performed the Kol- Table 2 and Supplementary Figure S2 show Spearman’s corre-
mogoroveSmirnov test to identify whether the data are not normally lation coefficients between carotid IMT changes (D IMT) and the
distributed. Because of their non-normal distribution, all continuous variables for glycemic indices and cardiovascular risks in enrolled
452 S.O. Song et al. / Atherosclerosis 225 (2012) 450e455

Table 1
Participants characteristics stratified by carotid IMT progression.

All patients (n ¼ 218) Carotid IMT P value

Group I (Non progressioin, n ¼ 141) Group II (Progression, n ¼ 77)


Age (years) 62.0 [53.0e68.0 ] 62.0 [53.0e68.0 ] 62.0 [53.8e69.3 ] .364
Male sex (n,%) 122 (56%) 74 (52.5%) 48 (62.3%) .160
DM duration (years) 6.0 [3.0e12.0 ] 6.0 [3.0e12.0 ] 7.0 [3.0e14.3 ] .519
Smoking (n, %) 107 (49.7%) 64 (45.4%) 43 (55.8%) .141
BMI (kg/m2) 25.3 [23.3e27.5 ] 25.3 [23.3e27.4 ] 24.3 [23.0e27.4 ] .440
Waist-to-hip ratio .92 [.88e.96 ] .92 [.87e.96 ] .93 [.89e.96 ] .156
Follow up (months) 24.0 [21.0e27.0 ] 24.0 [21.0 27.0 ] 24.0 [21.0 27.0 ] .668
SBP (mmHg) 126.0 [119.0e138.0 ] 125 .0 [118.0e137.0 ] 129.0 [119.0e142.3 ] .032
DBP (mmHg) 73.0 [68.0e80.5 ] 73.0 [67.0e80.0 ] 73.0 [68.0e82.0 ] .798
baseline A1c (%) 7.00 [6.40e7.95 ] 6.80 [6.40e7.83 ] 7.20 [6.60e8.10 ] .157
baseline GA (%) 14.6 [16.9e20.9 ] 16.3 [14.2e20.5 ] 18.1 [15.7e23.4 ] .016
baseline GA/A1c 2.17 [2.45e2.76 ] 2.40 [2.14e2.69 ] 2.48 [2.30e2.91 ] .013
Mean A1c 6.95 [6.50e7.58 ] 6.90 [6.45e7.46 ] 7.08 [6.68e7.87 ] .054
Mean GA 16.4 [14.2e19.7 ] 15.8 [14.1e18.6 ] 17.5 [14.8e21.7 ] .011
Mean GA/A1c 2.39 [2.14e2.58 ] 2.13 [2.34e2.54 ] 2.19 [2.49e2.85 ] .019
baseline mean IMT .72 [.63e.83 ] .72 [.64e.83 ] .72 [.62e.85 ] .961
baseline max IMT .87 [.74e1.03 ] .87 [.74e1.03 ] .86 [.73e1.05 ] .793
follow-up mean IMT .71 [.61e.83 ] .67 [.60e.78 ] .76 [.67e.91 ] <.001
follow-up max IMT .70 [.81e.96 ] .77 [.67e.89 ] .90 [.77e1.13 ] <.001
D mean IMT .01 [.05e.03 ] .04 [.07 w .01 ] .04 [.01e.08 ] <.001
D max IMT .04 [.11e.01 ] .08 [.13 w .05 ] .03 [.00e.11 ] <.001
Creatinine .90 [.78e1.09 ] .89 [.76e1.02 ] .96 [.80e1.18 ] .002
eGFR 82.0 [70.9e93.2 ] 83.9 [73.1e95.8 ] 80.9 [64.2e89.5 ] .004
Total protein (g/dL) 7.10 [6.80e7.40 ] 7.10 [6.78e7.40 ] 7.20 [6.80e7.50 ] .313
Albumin (g/dL) 4.50 [4.30e4.80 ] 4.50 [4.30e4.80 ] 4.50 [4.30e4.80 ] .514
AST 21.0 [17.0e26.0 ] 20.0 [16.8e25.3 ] 21.5 [16.8e28.8 ] .739
ALT 22.0 [15.0e32.0 ] 21.0 [15.0e30.0 ] 23.0 [16.0e36.0 ] .420
Total cholesterol (mg/dL) 166.0 [144.0e196.0 ] 166.0 [141.8e199.8 ] 165.0 [146.8e188.5 ] .814
TG (mg/dL) 124.0 [37.0e176.0 ] 121.5 [86.0e174.5 ] 124.5 [96.5e179.3 ] .751
HDL (mg/dL) 43.0 [37.0e54.5 ] 43.5 [37.0e54.3 ] 43.0 [35.8e54.3 ] .308
LDL (mg/dL) 91.4 [71.6e117.0 ] 91.5 [69.6e118.9 ] 91.2 [73.8e111.7 ] .879

Data presented as median [IQR] for continuous variables with non-normal distribution and n (%) for categorical variables.

subjects. In all subjects, all glycemic indices (baseline and mean GA/A1c, and mean A1c, GA and GA/A1c values from this study and
A1c, GA and GA/A1c, respectively) were significantly but weakly conventional risk variables were entered as independent variables
associated with D IMT. The baseline GA (p ¼ 0.02) and GA/A1c ratio (Table 3). In this analysis, we used two statistical models with
(p ¼ 0.04) showed significant association with D IMT in group II, but different independent variables. In model 1, we adopted SBP, eGFR,
not in group I (p ¼ 0.426, .709, respectively). However, baseline A1c and the glycemic indices. To avoid confounding factors and to
and mean A1c did not correlate with D IMT, and, mean GA and GA/ investigate the effects of serum GA, A1c and GA/A1c ratio on IMT
A1c across the study period showed only a weakly associated progression, we controlled for SBP and eGFR. Baseline GA/A1c (OR
tendency (p ¼ 0.098, and .064, respectively) in group II. 1.741, p ¼ 0.024), mean GA (OR 1.067, p ¼ 0.010), and mean GA/A1c
ratio (OR 2.255, p ¼ 0.010) were independently correlated with IMT
4.1. Factors associated with progression of carotid artery IMT progression. In model 2, we included classical atherosclerosis risk
factors such as age, gender, BMI, and LDL cholesterol in addition to
For the multivariate regression analysis, IMT change (D IMT) the variables in included in model 1. Baseline GA/A1c ratio, mean
(mm) was used as a dependent variable, and baseline A1c, GA, and GA, and mean GA/A1c ratio were independently correlated with

Fig. 1. (A) correlation between A1c and GA in each group; all patients: R ¼ .744, p < 0.001 Group I: R ¼ .719, p < 0.001 Group II: R ¼ .779, p < 0.001.(B). Correlation between mean
A1c and mean GA in each group; all patients: R ¼ .767, p < 0.001 Group I: R ¼ .749, p < 0.001 Group II: R ¼ .783, p < 0.001.
S.O. Song et al. / Atherosclerosis 225 (2012) 450e455 453

Fig. 2. Glycemic indices between two groups with or without carotid IMT progression.

D IMT in both models. Of the glycemic indices, a high GA level and physiological aspects of the GA/A1c ratio in type 2 diabetic subjects,
a high GA/A1c ratio were correlated with D IMT after multivariate the observation that the GA/A1c ratio increases more rapidly than
adjustment for other risk factors in both models 1 and 2 (Table 3 the A1c level may be attributed to more marked increases in GA
and Supplementary Table S1). levels relative to A1c levels and to certain diabetic conditions such
as poorly controlled glucose and insulin deficient status [3,15]. The
4.2. Area under curve of glycemic indices for predicting carotid IMT more rapid increase in GA relative to A1c reflects the fact that the
progression rate of non-enzymatic glycation of proteins in the circulation is
approximately nine times faster than that of human hemoglobin,
We calculated the AUC of glycemic indices for D IMT prediction which resides in erythrocytes, and the glycation of albumin is
(Fig. 3). Except for baseline A1c (p ¼ 0.157) and mean A1c across the therefore approximately 10-fold greater than that of hemoglobin
study period (p ¼ 0.053), the AUCs for other glycemic indices were [14,16,17]. However, the practical implications of an increased or
statistically significant. Among baseline glycemic indices, baseline decreased GA/A1c ratio with respect to the development of diabetic
GA/A1c ratio presented the largest AUC for D IMT (p ¼ 0.010, vascular complications have not been well investigated.
AUC ¼ .618). In respect to average glycemic values over study Based on the knowledge that increased serum GA is associated
period, mean GA had the largest AUC for D IMT (p ¼ 0.011, with diabetic vascular complications [13], we hypothesized that
AUC ¼ .621). The cut-off values of baseline GA, GA/A1c ratio, and increased GA levels might be associated with progression of
A1c were 15.5%, 2.274, and 6.9%, respectively. In addition, cut-off atherosclerosis. To test this hypothesis, we adopted the GA/A1c
values of mean GA, GA/A1c ratio, and A1c over the study period ratio used for adjusting conventional glucose levels and measured
were 16.3%, 2.47, and 6.81% respectively. progressive changes of carotid IMT (D IMT) as an end-point of
atherosclerosis. We investigated the following: (1) variables that
5. Discussion differ between patients with progression and non-progression of
carotid IMT; (2) whether conventional CVD risk factors and glyce-
In diabetic subjects, exposure of circulating proteins to hyper- mic indices measured by the standard glycemic index of A1c, gly-
glycemia is inevitable, resulting in protein glycation. Among such cated albumin, and GA/A1c ratio are associated with the
glycated proteins, there is growing evidence supporting the progression of carotid IMT; and (3) which variables could predict
atherogenic effects of GA, a precursor of AGEs, in diabetic subjects the progression of carotid IMT.
[2,4,5,6,11,12,13]. Among glycation indices for glucose monitoring, We compared glucose metabolism parameters and cardiovas-
the current golden standard is hemoglobin A1c. However, some cular risk variables according to carotid IMT progression and non-
discrepancies exist between the levels of A1c and glycated albumin progression. The patients with carotid IMT progression showed
(GA) in certain pathologic disease states including anemia, liver increased SBP, increased baseline and mean GA and GA/A1c ratio,
cirrhosis, and nephritic syndrome [14]. With respect to the and reduced renal function. Although the levels of A1c were not
significantly different between the groups, the levels of GA and
ratio of GA/A1c were significantly different.
Table 2 With respect to independent correlates of carotid IMT
Associations of cardiovascular risk factors with carotid IMT progression.
progression, a positive D IMT was associated with all glycemic
All patients Group I Group II indices in all patients, although baseline A1c only tended to be
b p-value b p-value b p-value associated with progression of carotid IMT in our study population.
Age .011 .874 .161 .057 .047 .686 Other CVD risk factors such as age, BMI, WH ratio, BP, and choles-
BMI .059 .383 .011 .898 .109 .344 terol showed no association with positive D IMT. The lack of asso-
WH ratio .025 .719 .013 .875 .227 .049 ciation between these CVD risk factors including A1c and a positive
SBP .100 .140 .037 .663 .191 .096 D IMT could be due to the characteristics of our study population. In
DBP .020 .764 .014 .873 .235 .040
baseline A1c .186 .006 .156 .064 .220 .055
our diabetic patients, well-controlled plasma glucose (median A1c,
baseline GA .206 .002 .068 .426 .342 .002 7.00%, IQR 6.40e7.95), blood pressure (median SBP, 126 mmHg, IQR
baseline GA/A1c .190 .006 .032 .709 .336 .004 119.0e138.0) and cholesterol (median LDL-C, 91.4 mg/dL, IQR 71.6e
mean A1c .184 .007 .130 .126 .140 .223 117.0), as well as the early stage of disease (median duration of
mean GA .172 .011 .025 .770 .190 .098
diabetes, 6.0 years, IQR 3.0e12.0) might explain the lack of asso-
mean GA/A1c .139 .040 .034 .685 .212 .064
creatinine .084 .218 .137 .106 .100 .388 ciation of these factors with the progression of carotid IMT.
eGFR .073 .283 .161 .056 .013 .910 With respect to predictors for carotid IMT progression, the
T.chol .039 .568 .002 .982 .342 .002 association between initial and mean GA levels and carotid IMT
TG .129 .058 .135 .112 .313 .006 progression was maintained after multivariate adjustment. Multi-
HDL .028 .682 .053 .533 .002 .985
variable logistic regression analysis investigating factors that affect
LDL .008 .903 .020 .811 .071 .540
the progression of carotid IMT showed that SBP, eGFR, baseline GA/
454 S.O. Song et al. / Atherosclerosis 225 (2012) 450e455

Table 3
Multivariable logistic regression models for associations of cardiovascular risk factors with carotid IMT progression.

A B C

OR 95% CI for OR P value OR 95% CI for OR P value OR 95% CI for OR P value


Model 1
SBP 1.022 1.003e1.042 .022 1.022 1.003e1.042 .021 1.024 1.004e1.043 .016
eGFR .979 .965e.994 .006 .978 .964e.993 .003 .980 .966e.995 .007
Baseline A1c
Baseline GA
Baseline GA/A1c 1.741 1.074e2.821 .024
Mean A1c
Mean GA 1.067 1.016e1.121 .010
Mean GA/A1c 2.255 1.216e4.183 .010
Model 2
Age .996 .966e1.027 .801 .998 .968e1.029 .910 .995 .965e1.026 .758
Gender .807 .442e1.472 .484 .809 .442e1.480 .492 .828 .452e1.516 .540
BMI .989 .925e1.058 .750 .992 .928e1.061 .820 .995 .930e1.065 .891
SBP 1.021 1.001e1.040 .036 1.021 1.002e1.040 .034 1.022 1.003e1.042 .027
eGFR .980 .963e.997 .018 .979 .962e.996 .015 .980 .963e.997 .021
LDL 1.001 .993e1.009 .792 1.001 .993e1.009 .833 1.001 .993e1.010 .736
Baseline A1c
Baseline GA
Baseline GA/A1c 1.680 1.013e2.788 .045
Mean A1c
Mean GA 1.063 1.010e1.119 .018
Mean GA/A1c 2.213 1.148e4.265 .018

A1c ratio, mean GA, and mean GA/A1c ratio were independently dysfunctional endogenous insulin secretion [6], clinical significance
correlated with carotid IMT progression in this study population. of GA/A1c ratio over GA or A1c alone are not only a value of adjusted
When we analyzed our data using only serum A1c as a glycemic conventional glucose levels but is also a better reflection of glucose
parameter rather than serum GA and GA/A1c ratio, no correlation fluctuation, postprandial hyperglycemia, and dysfunctional insulin
with IMT progression was found with any of the analytic models. In secretory function [15].
this regard, we suggest that GA level might be a more powerful In understanding the pathogenic effects of glycated albumin
parameter in predicting carotid IMT progression than conventional (GA) on diabetic vascular complications, it is necessary to consider
golden standard glycemic index of hemoglobin A1c. Regarding to that GA results from early-stage Amadori-modified reaction prod-
significance of GA/A1c ratio compared to GA or A1c alone, simple ucts formed from Schiff’s base adducts [14], and is regarded as
discrepancies between the levels of GA and A1c are attributed to a precursor of AGEs [5]. In addition, albumin, synthesized mainly in
certain pathologic conditions, such as chronic renal failure and the liver, constitutes 50e60% of the total protein in the plasma of
anemia, or to rapid glucose changes such as developing time of healthy subjects. Physiologically, it is a pivotal antioxidant in
fulminant type 1 diabetes and periods of intensified glucose human serum [20]. Therefore, glycation and accompanying oxida-
lowering medication. Glucose excursion is a well-known factor for tion of albumin leads to a loss of antioxidant activity and generates
atherosclerosis, especially carotid intima-media thickness [18,19]. an atherogenic protein in patients with diabetes [6]. This could
Because GA, even among similar A1c levels, better reflects post- explain why a higher level of GA is related to vascular complications
prandial hyperglycemia, which is mainly caused by inadequate or in diabetes [5,12].

Fig. 3. ROC Curve of each glycemic indices predicting IMT progression.


S.O. Song et al. / Atherosclerosis 225 (2012) 450e455 455

Our study has some limitations. First of all, this was a retro- [3] Lee EY, Lee BW, Kim D, et al. Glycated albumin is a useful glycation index for
monitoring fluctuating and poorly controlled type 2 diabetic patients. Acta
spective study conducted at a single center. Second, although other
Diabetol 2011;48:167e72.
variables such as medications taken by the study subjects and [4] Pu LJ, Lu L, Xu XW, et al. Value of serum glycated albumin and high-sensitivity
smoking habits might affect the levels of GA, we did not analyze C-reactive protein levels in the prediction of presence of coronary artery
their influence on GA because our attention was focused on the disease in patients with type 2 diabetes. Cardiovasc Diabetol 2006;5:27.
[5] Rodino-Janeiro BK, Gonzalez-Peteiro M, Ucieda-Somoza R, et al. Glycated
clinical relevance of a higher GA/A1c ratio to atherosclerosis. albumin, a precursor of advanced glycation end-products, up-regulates
However, previous studies have demonstrated that smoking, sta- NADPH oxidase and enhances oxidative stress in human endothelial cells:
tins, and oral hypoglycemic agents are associated with serum levels molecular correlate of diabetic vasculopathy. Diabetes Metab Res Rev 2010;
26:550e8.
of GA [6,14,21,22,23,24] and this should be considered in future [6] Kim KJ, Lee BW. The roles of glycated albumin as intermediate glycation index
studies. and pathogenic protein. Diabetes Metab J 2012;36:98e107.
Based on the findings of the present study, we conclude that [7] Moon JH, Chae MK, Kim KJ, et al. Decreased endothelial progenitor cells and
increased serum glycated albumin are independently correlated with plaque-
serum GA has two main clinical implications in diabetic subjects: forming carotid artery atherosclerosis in type 2 diabetes patients without
first, similar to previous reports [3,13,15,25], GA is strongly corre- documented ischemic disease. Circ J 2012;76(9):2273e9.
lated with A1c (r ¼ .744, p < 0.001) which confers utility for the [8] Kim ST, Kim BJ, Lim DM, et al. Basal C-peptide level as a surrogate marker of
subclinical atherosclerosis in type 2 diabetic patients. Diabetes Metab J 2011;
glycemic index; second, GA might be an atherogenic protein in 35:41e9.
diabetic subjects because our study demonstrates a significant [9] Kim SK, Park SW, Kim SH, et al. Visceral fat amount is associated with carotid
association between progression of IMT (D IMT) and glycated atherosclerosis even in type 2 diabetic men with a normal waist circumfer-
ence. Int J Obes (Lond) 2009;33:131e5.
albumin-related parameters. This results of this experiment are in
[10] Touboul PJ, Hennerici MG, Meairs S, et al. Mannheim carotid intima-media
accord with the earlier one [13]. We postulate that new parameters thickness consensus (2004-2006). An update on behalf of the advisory
or equations to predict CVD using serum GA and GA/A1c ratio might board of the 3rd and 4th watching the risk symposium, 13th and 15th
be developed in future studies with a larger study population. European stroke conferences, Mannheim, Germany, 2004, and Brussels,
Belgium, 2006. Cerebrovasc Dis 2007;23:75e80.
[11] Selvin E, Francis LM, Ballantyne CM, et al. Nontraditional markers of glycemia:
associations with microvascular conditions. Diabetes Care 2011;34:960e7.
Funding [12] Cohen MP, Shea E, Chen S, et al. Glycated albumin increases oxidative stress,
activates NF-kappa B and extracellular signal-regulated kinase (ERK), and
This study was supported by a faculty research grant from stimulates ERK-dependent transforming growth factor-beta 1 production in
macrophage RAW cells. J Lab Clin Med 2003;141:242e9.
Yonsei University College of Medicine, 2012 (6-2012-0009).
[13] Furusyo N, Koga T, Ai M, et al. Plasma glycated albumin level and athero-
sclerosis: results from the Kyushu and Okinawa population study (KOPS). Int J
Cardiol 2012 May 31 [Epub ahead of print].
Acknowledgments [14] Rondeau P, Bourdon E. The glycation of albumin: structural and functional
impacts. Biochimie 2011;93:645e58.
Statistical analyses in this study were supported by Jieun Kim [15] Kim D, Kim KJ, Huh JH, et al. The ratio of glycated albumin to glycated hae-
moglobin correlates with insulin secretory function. Clin Endocrinol (Oxf)
(Basic Research Assistant Professor) and Min Woong Kang 2011 Dec 9 [Epub ahead of print].
(Research Assistant) at Biostatistics Collaboration Unit, Yonsei [16] Garlick RL, Mazer JS. The principal site of nonenzymatic glycosylation of
University College of Medicine, Seoul, Korea. human serum albumin in vivo. J Biol Chem 1983;258:6142e6.
[17] Iberg N, Fluckiger R. Nonenzymatic glycosylation of albumin in vivo. Identi-
Dong-Su Jang, B,A., (Research Assistant, Department of
fication of multiple glycosylated sites. J Biol Chem 1986;261:13542e5.
Anatomy, Yonsei University College of Medicine, Seoul, Korea) [18] Chen XM, Zhang Y, Shen XP, et al. Correlation between glucose fluctuations
supplied figures for our requests. and carotid intima-media thickness in type 2 diabetes. Diabetes Res Clin Pract
The authors greatly appreciate the expert technical assistance 2010;90:95e9.
[19] Hu Y, Liu W, Huang R, et al. Postchallenge plasma glucose excursions, carotid
and help of them. intima-media thickness, and risk factors for atherosclerosis in Chinese pop-
ulation with type 2 diabetes. Atherosclerosis 2010;210:302e6.
[20] Halliwell B. Albuminean important extracellular antioxidant? Biochem
Appendix A. Supplementary material Pharmacol 1988;37:569e71.
[21] Koga M, Saito H, Mukai M, et al. Serum glycated albumin levels are influenced
Supplementary data associated with this article can be found, in by smoking status, independent of plasma glucose levels. Acta Diabetol 2009;
46:141e4.
the online version, at http://dx.doi.org/10.1016/j.atherosclerosis. [22] Thongtang N, Ai M, Otokozawa S, et al. Effects of maximal atorvastatin and
2012.09.005. rosuvastatin treatment on markers of glucose homeostasis and inflammation.
Am J Cardiol 2011;107:387e92.
[23] Bao YQ, Zhou M, Zhou J, et al. Relationship between serum osteocalcin and
References glycaemic variability in Type 2 diabetes. Clin Exp Pharmacol Physiol 2011;38:
50e4.
[1] Jeong IK, King GL. New perspectives on diabetic vascular complications: the [24] Tang SC, Leung JC, Chan LY, et al. Renoprotection by rosiglitazone in accel-
loss of endogenous protective factors induced by hyperglycemia. Diabetes erated type 2 diabetic nephropathy: role of STAT1 inhibition and nephrin
Metab J 2011;35:8e11. restoration. Am J Nephrol 2010;32:145e55.
[2] Lu L, Pu LJ, Zhang Q, et al. Increased glycated albumin and decreased esRAGE [25] Furusyo N, Koga T, Ai M, et al. Utility of glycated albumin for the diagnosis of
levels are related to angiographic severity and extent of coronary artery diabetes mellitus in a Japanese population study: results from the Kyushu and
disease in patients with type 2 diabetes. Atherosclerosis 2009;206:540e5. Okinawa population study (KOPS). Diabetologia 2011;54:3028e36.

Potrebbero piacerti anche