Nephrol Dial Transplant (2010) 25: 1218–1225
doi: 10.1093/ndt/gfp558
Advance Access publication 21 October 2009
Pre-diabetes and arterial stiffness in uraemic patients
Mads Hornum1, Peter Clausen1, Jesper Kjaergaard2, Jesper Melchior Hansen3,
Elisabeth R. Mathiesen4 and Bo Feldt-Rasmussen1
1
Department of Nephrology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark, 2Department of Cardiology,
Rigshospitalet, University of Copenhagen, Copenhagen, Denmark, 3Department of Nephrology, Herlev Hospital, University of
Copenhagen, Copenhagen, Denmark and 4Department of Endocrinology, Rigshospitalet, University of Copenhagen, Copenhagen,
Denmark
Correspondence and offprint requests to: Mads Hornum; E-mail: [email protected]
Abstract
Background. In order to address factors of relevance for
new onset diabetes mellitus and cardiovascular disease after kidney transplantation, we investigated the presence of
pre-diabetes, arterial stiffness and endothelial dysfunction
in patients with end-stage renal disease (ESRD) accepted
for kidney transplantation.
Methods. Pre-diabetes and an insulin sensitivity index
were estimated by an oral glucose tolerance test in 66 consecutive uraemic patients, without diabetes, being on the
waiting list for the first renal transplantation. Mean age
was 43 ± 13 years. Duration of ESRD was 32 ± 27 months.
A control group consisted of 14 healthy subjects.
Arterial stiffness was measured by aorta pulse wave velocity
(PWV) and aorta augmentation index (AIX). Endothelial
function was evaluated by flow-mediated vasodilatation
(FMD) and plasma concentrations of von Willebrand factor
antigen (vWF). Mean arterial blood pressure (MAP) was
measured in supine resting position.
Results. Twenty-seven uraemic patients (41%) had pre-diabetes (IFG+IGT), and 39 had normal glucose tolerance.
The uraemic patients were more insulin resistant with lower insulin sensitivity index compared to healthy controls
(6.1 ± 3 vs. 15 ± 7, P < 0.0001) but with no difference
between patients with and without pre-diabetes. HbA1c
and fasting plasma glucose was comparable in uraemic patients with and without pre-diabetes.
PWV was higher in pre-diabetic compared to normoglycaemic uraemic patients (9.1 ± 3 vs. 7.3 ± 2 m/s, P =
0.03) and healthy controls (9.1 ± 3 vs. 6.7 ± 1, P =
0.01), while AIX did not differ (24.9 ± 13 vs. 23.2 ± 12
vs. 17 ± 16, P = NS). Presence of pre-diabetes was positively associated to PWV in a univariate analysis. Multivariable analysis revealed age and MAP as independent
predictors of PWV in uraemic patients. FMD and vWF
were impaired in uraemic patients compared to healthy
controls (3 ± 4 vs. 7 ± 3, P = 0.007 and 1.8 ± 0.7 vs.
0.96 ± 0.3 kIU/L, P = 0.0002, respectively) but with no
difference between the two groups of uraemic patients.
In conclusion, a high prevalence of pre-diabetes, impaired
insulin resistance, increased arterial stiffness of aorta as
well as impaired augmentation index and vasodilatation
was demonstrated in uraemic patients prior to kidney
transplantation. Increased arterial stiffness of aorta and
augmentation index were independently associated with
age and blood pressure.
Keywords: arterial stiffness; augmentation index; insulin resistance; prediabetes; pulse wave velocity
Introduction
New onset diabetes mellitus after solid organ transplantation is an important clinical challenge associated to
increased risk of major cardiovascular events after transplantation [1,2]. Previous studies have focused on metabolic changes as they appear after transplantation, and the
presence of pre-diabetes at baseline before transplantation
has only been sparsely investigated [3,4]. However, especially in kidney transplant recipients, studies before transplantation are highly relevant since uraemic patients are
characterized by increased insulin resistance. Pre-diabetes
may therefore be prevalent in patients with end-stage renal disease (ESRD) [5–8] and potentially associated with
vascular damage.
Arterial stiffness can be estimated by aorta pulse wave
velocity (PWV) [9], which reflects damage of the elastic
tissue of aorta. An increased PWV has been shown to be
an integrated index of vascular risk factors [10]. It is associated with increased mortality in ESRD [3,11,12] and
is associated with hypertrophy of the arterial walls [13]
and arterial calcifications [14]. Arteries are stiffer in patients with type 1 and 2 diabetes with different degrees
of diabetic and cardiovascular complications [15–17],
and a potential association of impaired glucose tolerance
or pre-diabetes with diminished arterial elasticity in pre-diabetes may exist. Thus, presence of pre-diabetes in ESRD
patients may also be associated with further increased arterial stiffness leading to increased risk of cardiovascular
events [18]. Similarly, increased pulse pressure is predic-
© The Author 2009. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please e-mail: [email protected]
Pre-diabetes and arterial stiffness in uraemia
tive of cardiovascular mortality in type 2 diabetes [19,20],
and augmentation of the pulse pressure is, at least in part,
secondary to increased stiffness of mainly the resistance
arteries. It can be measured by applanation tonometry using the pressure waveform at the radial artery. Mathematically, it is transformed into the pressure waveforms in
ascending aorta representing the true workload of the left
ventricle [21].
Endothelialfunctionmeasured byendothelium-dependent
dilatory response of the brachial artery has also been found
to be impaired in both uraemia [22] and diabetes [23] and
may reflect presence of vascular disease [24] or early vascular damage.
A relationship between presence of pre-diabetes and increased arterial stiffness and endothelial dysfunction may
therefore be present in the uraemic population. We studied
the presence of pre-diabetes, arterial stiffness and endothelial dysfunction in a group of non-diabetic patients with
ESRD accepted for kidney transplantation. The aim was
to determine the prevalence of pre-diabetes in uraemic
patients prior to transplantation and to investigate the associations between pre-diabetes, arterial stiffness and endothelial dysfunction.
Materials and methods
Patients
All pre-dialysis, haemodialysis and peritoneal dialysis patients, aged between 18–65 years, accepted for and awaiting their first kidney transplantation were screened. Exclusion criteria were medical history of diabetes
or diabetes diagnosed during oral glucose tolerance test (OGTT). The
screening was conducted in the period between January 2006 and March
2008 at the regional transplantation centres at Rigshospitalet and Herlev
Hospital. Out of 510 patients from the Scandiatransplant waiting list for
kidney transplantation, 140 patients who fulfilled the inclusion criteria
were invited to participate. Seventy-three patients (52%), mostly Nordic
Caucasians, accepted participation in the study, of which 7 (5%) were
found to have diabetes by careful review of the patients' record (n = 4)
or OGTT (n = 3), which left 66 uraemic patients without diabetes to be
studied. The main reasons for non-attendance were long travelling time to
the laboratory or inability to use extra days at the clinic besides the time
already used for dialysis.
The participants were asked about previous diagnosis of diabetes,
acute myocardial infarction and stroke, and the medical records were carefully reviewed for these diagnoses. A control group consisting of 14
healthy age, body mass index (BMI) and sex matched subjects was recruited from public announcing.
The regional ethical committee (# KF 01279825) and The Data Protection Agency (#2006-41-5640) approved the study. All participants gave
their informed written consent.
Study procedure
All participants were examined after an overnight fast including coffee,
tobacco and exercise abstinence for 10 h. Usual antihypertensive medication was allowed in the morning.
Haemodialysis patients were examined between the days of haemodialysis. Peritoneal dialysis patients had the peritoneal fluid drained in the
morning at 6 p.m. The examination began between 8 and 11 p.m. After
10 min of rest in the supine resting position, blood pressure was measured
in triplicate from the arm opposite to a fistula or dialysis catheter. Mean
arterial blood pressure (MAP) was based on the mean of these three measurements. Fasting blood samples were drawn for the determination of
glucose, HbA1C, homocystein, von Willebrand factor antigen (vWF),
haematological factors, lipids, urate, C-reactive protein and parathyroidea
hormone (PTH) (Table 3).
1219
All blood samples were taken from an antecubital vein. Plasma glucose
concentrations were analysed by the glucose-hexokinase method (Glucoquant®, Roche Diagnostics GmbH, D-68298 Mannheim, Germany), and
insulin and PTH were measured using enzyme-linked immunosorbent assay kits (Elecsys, Roche Diagnostics GmbH, D-68298 Mannheim, Germany). All assays were automated and performed on a Cobas Fara
robot (Roche Diagnostics GmbH, Mannheim, Germany). Standard laboratory methods were applied for the analysis of the other samples. The
examiners were blinded regarding the actual clinical and metabolic status
of the patient at the time of examination, and all clinical information was
analysed and described after data collection was completed.
Oral glucose tolerance test
A 75-g oral glucose tolerance test was performed according to WHO/
ADA 2007 criteria [25]. The prevalence of normal glucose tolerance
(NGT: 2-h post load glucose <7.8 mmol/l), impaired fasting glucose
(IFG: fasting plasma glucose between 5.6 and 6.9 mmol/l) and impaired
glucose tolerance (IGT: 2-h post load glucose between 7.8–11.1) as well
as pre-diabetes (IFG+IGT) was determined. Plasma concentrations of
glucose and insulin were measured at times −30, −15, 0, 30, 60, 90
and 120 min. According to Matsuda et al., an insulin sensitivity index
(ISI composite) was calculated as 10 000/square root of [fasting glucose ×
fasting insulin] × [mean glucose × mean insulin during OGTT]. This index
is highly correlated with the rate of whole-body glucose disposal during
a euglycemic insulin clamp in patients with varying degrees of glucose
tolerance [26].
The area under the curve (AUC) for insulin and glucose during the
OGTT was calculated using the trapezoid rule. These variables were
implemented in the insulin secretion index; SecrAUC = AUCIns/AUCGlu
[27, 28].
Pulse wave analysis and velocity
Pulse wave velocity of the carotid-femoral pulse wave was recorded sequentially at the carotid and femoral arteries using applanation tonometry
(SphygmoCor®; AtCor Medical, Sydney, NSW, Australia) performed by
the same observer (MH). Using simultaneous three-lead electrocardiogram
recording of the R wave as the reference frame and calculating the time
delay using a foot-of-the-wave method (intersecting tangent), the PWV
was calculated by the system software. The distance between the carotid
artery recording site and the sternal notch was subtracted from the distance
from the sternal notch to the palpable femoral pulse. PWV was calculated as
ratio between the distances travelled by the pulse wave and pulse transmission time.
The pulse wave was recorded from the radial artery of the right arm or,
in the case of a fistula, from the contralateral arm with a high-fidelity
micromanometer (SPC-301; Millar Instruments, Houston, TX, USA). A
model of the central pressure waveform was synthesized by the Sphygmocor software using a validated transfer function as previously described
[9]. The pulse wave analysis (PWA) was used to determine aortic augmentation as the difference between the second (caused by wave reflection)
and the first systolic peak (caused by ventricular ejection). The augmentation index (AIX) serves as an integrated parameter for the cardiovascular function in the arteries and is calculated from the augmentation (ΔP)
and the central pulse pressure (PP) computed by the software from brachial systolic and diastolic pressure, (AIX = (ΔP/PP) × 100). An average
of two consecutive readings, each consisting of at least 20 sequentially
recorded waveforms, was used for the analyses. AIX was adjusted for
heart rate. Time to the foot of the reflected wave (Tr) was identified
and represents the composite travel time of the pulse wave to the periphery and its return. It is enhanced by increased pulse wave velocity and is a
further index of conduit artery stiffness. Subendocardial variability ratio
(SEVR) is the ratio of diastolic pressure time interval (AD)/systolic pressure time interval (AS) [29] and was also determined. SEVR is associated
with the pressure and time for coronary perfusion, and this has, in patients
with known coronary artery disease, been found to indicate aggravating
subendocardial ischaemia [30] due mainly to reduction in diastolic perfusion time. The reproducibility of PWA and PWV for two repeat scans
of 10 uraemic patients from the present study was analysed using the
Bland-Altman method. The mean (±SD) intra-observer difference was
0.1 ± 0.7 m/s for PWV, 1.7 ± 5.5% for AIX and −1.0 ± 8.8% for SEVR,
which is in accordance with other groups [31,32].
1220
M. Hornum et al.
Table 1. Clinical characteristic of 66 uraemic patients with and without pre-diabetes and 14 healthy controls
Age (years)
Sex (M/F)
BMI (kg/m2)
Waist–hip ratio
Ever smoking, n (%)
CVD, n (%)
Previous MI
Apoplexia
Dialysis status
Hemodialysis (n)
CAPD (n)
Pre-dialysis (n)
ESRD (months)
Diagnoses (n)
Glomerulonephritis
Hypertension
PKD
Vasculitis
Other/unknown
Antihypertensive treatment (n)
Mean number of drugs
AII-antagonist
ACE-Inhibitor
Ca-antagonist
Diuretic
Beta-blocker
Alfa-blocker
Normal glucose tolerance
Uraemic patients
N = 39
Pre-diabetes
Uraemic patients
N = 27
Healthy controls
N = 14
41 ± 12
31/8
25 ± 4
0.93 ± 0.09
27/39 (69)
46 ± 12
14/13*
25 ± 4
0.90 ± 0.07
12/ 27 (44)
39 ± 11
9/5
24 ± 3
0.81 ± 0.08§§§
9/14 (64)
0
4 (10)
1 (4)
1( 4)
–
25 (64)
11 (28)
3 (8)
31 ± 28
13 (48)
12 (44)
2 (7)
33 ± 26
–
–
–
–
15 (39)
9 (23)
4 (10)
1 (3)
10 (26)
8
4
5
1
9
(30)
(15)
(19)
(4)
(33)
–
–
–
–
–
2 (0–6)
12 (31)
12 (31)
22 (56)
20 (51)
18 (46)
2 (5)
2 (0–6)
13 (48)
5 (19)
14 (52)
18 (67)
13 (45)
3 (11)
–
–
–
–
–
–
Data are presented as mean ± SD, numbers (n) and percent (%). CVD cardiovascular disease, CAPD continous ambulatory peritoneal dialysis, ESRD endstage renal disease, PKD polycystic kidney disease. Wilcoxon rank sum test or unpaired t-test used where appropriate to test: Normal glucose tolerance
patients vs. pre-diabetes patients; *P < 0.05. Uraemic patients vs. healthy controls; §§§P < 0.0005.
Flow-mediated vasodilatation and nitroglycerin-induced dilatation
An Acuson 128XP/10™ ultrasound system with extended frequencyimaging option and a 7-MHz linear transducer were used to measure
the brachial artery dilatory responses to increased flow and to nitroglycerin (NTG) based on previous published methodology by our group and
others [33,34].
FMD is dependent on the endothelial function [35], while nitroglycerininduced dilatation (NID) is the maximal vasodilatory capacity independent
of the endothelial function.
The ultrasound system was connected to a video recorder and to a
computer equipped with a frame grabber and artificial neural network
wall detection software (vessel image analysis (VIA)) developed by Sidhu
et al. [33]. The VIA software automatically detects and tracks the anterior
and posterior walls within a user-defined region of interest. Arterial diameter was measured by VIA at baseline (mean over 15 s) and continuously 5 min after release of a pneumatic tourniquet, which had been
inflated for 4.5 min to a pressure of 300 mmHg distal to the segment
of the artery being scanned. After 10 min of rest, 400μg NTG (Nitrolingual spray, Pohl-Boskamp, Hohenlockstedt, Germany) was administrated
sublingually and arterial diameter again measured continuously for 10
min. FMD (defined as mean arterial diameter in percentage of baseline
diameter over a 15-s period at 55 s after tourniquet deflation) was automatically calculated. NID was calculated similarly as mean arterial diameter in percentage of baseline diameter over a 15-s period at 5 min after
NTG administration.
Doppler-derived flow was measured at videotape using a pulsedwave Doppler signal at a 70° angle to the vessel with the range gate
(1.5mm) in the centre of the artery at baseline (mean of four measurements) and at the maximum within the first 15 s after tourniquet release.
Blood flow was calculated by multiplying the velocity time integral of
the Doppler flow signal (corrected for angle) by the heart rate and the
vessel cross-sectional area and percentage flow increase subsequently
calculated.
External ultrasound measurements of FMD and NID have previously
been shown to provide accurate and reproducible results [24,36] Coefficients of variation of the diameter changes expressed as diameter in percentage of the baseline scan diameter (100%) in the present study was
4.4% and 4.1% for FMD and NID. This estimate was based on eight scans
of the same subject from the present study population (within-patient variability). The measurements were performed on separate days within 2
weeks by the same investigator, and the variation was similar to previous
reports [34,37].
Data regarding diagnosis of kidney disease, dialysis modality, duration
of ESRD, antihypertensive treatment and previous myocardial infarction
(MI) or apoplexia were obtained anamnestic and from the medical record.
Smoking history was obtained by questioning and given as ever smoking
(present and previous) or not. Waist–hip ratio and weight and height were
also measured.
Statistics
Data analyses were performed using Statistical Analysis Software (SAS®)
version 9.1. Unless specified otherwise, continuous data are described as
mean ± standard deviation (SD) or median and inter-quartile range (IQR)
for normal and skewed distributions, respectively. Group comparisons of
continuous data were performed using unpaired t-test for normally distributed data or non-parametric Wilcoxon rank sum test. Chi-square or Fisher's exact tests were used for group comparisons between categorical data
where appropriate. Correlation analysis was performed using Spearman
rank analysis. Univariate and multivariable analyses were performed with
PWV, AIX, FMD and NID as dependent variables. Based on the close
relationship between pre-diabetes and AUC glucose, we have chosen to
perform a multi regression analysis whereby the dependent variable is adjusted for age (years), gender (F/M), BMI (kg/m2), heart rate (beats per
minute) and MAP (mmHg) and then either pre-diabetes (Y/N), ISI or
AUC glucose (mM × 120 min) was added.
A P value of <0.05 was used to determine significance.
Pre-diabetes and arterial stiffness in uraemia
1221
Table 2. Glycaemic, inflammatory, endothelial and uraemic variables in 66 uraemic patients with and without pre-diabetes and 14 healthy controls
Glycaemic
Fasting p-glucose (mmol/L)
HbA1C (%)
P-glucose at 2 h (mmol/L)
Insulin sensitivity index (ISI-composite)
AUC glucose (mmol/L·120 min)
AUC insulin (pmol/L·120 min)
Insulin secretion index
Inflammatory and endothelial
Von Willebrand factor antigen (kIU/L)
C-reactive protein (mg/L)
P-homocystein (μmol/L)
P-urate (mmol/L)
Uraemic and lipids
P-creatinine (μmol/L)
P-carbamid (mmol/L)
P-haemoglobin (mmol/L)
P-PTH (pmol/L)
P-total cholesterol (mmol/L)
S-LDL-cholesterol (mmol/L)
S-HDL-cholesterol (mmol/L)
P-triglyceride (mmol/L)
Normal glucose tolerance
Uraemic patients
N = 39
Pre-diabetes
Uraemic patients
N = 27
Healthy controls
N = 14
5.1 ± 0.4
5.2 ± 0.3
6.7 ± 1.0
7.3 ± 3.8
7.7 ± 1.1
290 ± 189
38 ± 22
5.2 ± 0.4
5.3 ± 0.4
8.8 ± 1.0***
6.1 ± 3.3
8.9 ± 1.2*
262 ± 141
30 ± 17
5.0 ± 0.3
5.2 ± 0.2
5.4 ± 1.1§§§
14.7 ± 7.0§§§
6.3 ± 1.2§§§
174 ± 107§
27 ± 14
1.8 ± 0.7
4±3
25 ± 13
0.35 ± 0.1
1.7 ± 0.6
5±8
28 ± 13
0.33 ± 0.1
0.96 ± 0.3§§§
1 ± 1 §§§
14 ± 5 §§§
0.28 ± 0.1§
874 ± 305
20 ± 7
7.6 ± 1
35 ± 32
4.7 ± 1.3
3.0 ± 1.2
1.4 ± 0.5
1.8 ± 1.2
889 ± 444
21 ± 9
7.4 ± 1
24 ± 20
4.7 ± 1.3
2.9 ± 1.1
1.4 ± 0.5
1.7 ± 1.3
69 ± 9 §§§
4.7 ± 1.2 §§§
8.5 ± 1§§§
3.9 ± 1 §§§
4.1 ± 0.7§
2.4 ± 0.7§
1.7 ± 0.6
0.9 ± 0.56 §
Data are presented as mean ± SD, percent (%) or median with range. Wilcoxon rank sum test or unpaired t-test used where appropriate to test: Normal
glucose tolerance patients vs. pre-diabetes patients; *P < 0.05, ***P < 0.0005. Uraemic patients vs. healthy controls; §P < 0.05, §§§P < 0.0005.
Results
Twenty-seven (41%) uraemic patients had pre-diabetes of
which 25 had IGT and two IFG. The remaining 39 were
normoglycaemic. Uraemic patients with pre-diabetes were
comparable to normoglycaemic patients with regard to the
causes of renal disease, modality of dialysis, duration of
ESRD, smoking, blood pressure or antihypertensive medication (Table 1). Cardiovascular events [previous myocardial infarction (MI) or apoploxia] were registered in four
(10%) of the normoglycaemic and two (7%) of the prediabetes patients (NS, Table 1).
Glycaemic data
Pre-diabetic uraemic patients had a higher 2-h glucose as
compared with patients with normal glucose tolerance
(Table 2), but fasting glucose and HbA1C were similar to
the normoglycaemic uraemic patients. All the age-, sexand BMI-matched healthy controls were normoglycaemic.
The ISI-composite value of the pre-diabetic uraemic patients was comparable to the normal glucose tolerant uraemic patients, whereas the uraemic patients had significantly
reduced ISI composite compared to healthy controls. The
insulin secretion index, SecrAUC tended to be lower in uraemic patients with pre-diabetes (P = 0.1) but comparable to
the healthy controls (Table 2).
Pulse wave velocity and analysis
Carotid-femoral PWV was significantly increased in the
pre-diabetic uraemic patients compared to the patients with
normal glucose tolerance (Figure 1). No difference in
PWVor AIX between haemodialysis and peritoneal dialysis
patients were found (7.9 ± 3 m/s vs 8.8 ± 3 m/s and 26 ±
13 vs 23 ± 12, P = NS), but a significantly higher AIX was
found in dialysis patients as compared to pre-dialysis patients (15 ± 11 vs 24 ± 13, P = 0,01). PWV was significantly positively related to age (P < 0.0001), MAP (P =
0.005), baseline brachial artery diameter (P = 0.024),
AUCGlu120 min (P = 0.01) and pre-diabetes (P = 0.026)
in a univariate analysis. Using multivariable analysis with
pre-diabetes, age, sex, height, heart rate, MAP and dialysis
status as variables, age (P < 0.0001) and MAP (P = 0.004)
were independently related to PWV, and the model accounted for 53% of the variability in PWV (Table 4).
Using AUCglu 120 or ISI instead of pre-diabetes did not result in any change in independent factors. Pulse wave analysis measured by augmentation index, Tr and SEVR revealed
no difference between uraemic patients with and without
pre-diabetes (Table 3). Augmentation index was positively
associated to MAP (P = 0.04) and female sex (P = 0.001)
and inversely associated to heart rate (P < 0.005) and height
(P = 0.0004) in a univariate analysis. Using multivariable
analysis with pre-diabetes, age, sex, height, heart rate and
dialysis status as variables, age (P = 0.0004), female sex
(P = 0.004), heart rate (P < 0.0001), height (P = 0.01), dialysis status (P = 0.01) and MAP (P = 0.002) were independently associated to augmentation index, and the model
explained 63% of the variability in augmentation index
(Table 4). Using AUCglu120 or ISI instead of pre-diabetes
did not result in any change in independent factors.
The baseline brachial artery diameter did not improve
the models, although it was associated with increased
PWV in univariate analysis (data not shown). Type or
number of antihypertensive drugs did not influence the
1222
M. Hornum et al.
Fig. 1. Pulse wave velocity in 66 uraemic patients with and without pre-diabetes and 14 healthy controls.
Table 3. Arterial stiffness, endothelial and cardial function variables in 66 uraemic patients with and without pre-diabetes and 14 healthy controls
Blood pressure
Systolic (mmHg)
Diastolic (mmHg)
Pulse pressure
MAP
Heart rate
Arterial stiffness
Aortic PWV (m/s)
SEVR (%)
Augmentation index (%)
Time to reflected wave, Tr (ms)
Flow mediated vasodilatation
Baseline diameter in brachial artery (mm)
Flow mediated dilatation (% of baseline)
Flow increase in brachial artery after ischaemia (% of baseline)
Nitroglycerin-induced dilatation (% of baseline)
Normal glucose tolerance
Uraemic patients
N = 39
Pre-diabetes
Uraemic patients
N = 27
Healthy controls
N = 14
140 ± 21
84 ± 14
56 ± 16
103 ± 15
70 ± 13
144 ± 23
85 ± 10
59 ± 17
104 ± 13
70 ± 10
118 ± 10§§§
73 ± 8§§§
45 ± 8§§§
88 ± 8§§§
61 ± 9§§
7.3 ± 2
155 ± 30
23 ± 13
144 ± 11
9.1 ± 3*
148 ± 29
25 ± 12
145 ± 14
6.7 ± 1
191 ± 28§§§
17 ± 16
156 ± 14§§
4.3 ± 0.8
4 ± 5 (−1–19)
470 ± 200
12 ± 6 (2–27)
4.3 ± 0.9
3 ± 4 (−6–12)
480 ± 300
12 ± 5 (3–23)
3.8 ± 0.7§
7 ± 3 (1–14)§§
670 ± 300§
19 ± 5 (12–28)§§§
Data are presented as mean ± SD and percent (%) or range. PWV: pulse wave velocity, SEVR: subendocardial variability ratio, Wilcoxon Rank Sum
Test or unpaired t-test used where appropriate to test: Normal Glucose Tolerance patients vs. Pre-diabetes patients; *P < 0.05. Uraemic patients vs.
Healthy controls; §P < 0.05, §§P < 0.005, §§§P < 0.0005.
augmentation index or pulse wave results, and it was evenly distributed between groups (Table 1).
Flow-mediated vasodilatation
FMD, NID and the increase in brachial blood flow after
ischaemia were impaired in uraemic patients as compared
to healthy controls (P < 0.005, P < 0.0005 and P < 0.05).
However, there was no difference in FMD, NID, baseline
arterial diameter or increase in brachial flow between urae-
mic patients with and without pre-diabetes (Table 3). FMD
was associated to height (P = 0.04) in univariate regression
analysis, but with FMD or NID as the dependent variable
in the multivariate analysis, no significant associations to
independent variables were disclosed.
Inflammatory, endothelial and uraemic markers
Plasma homocystein, vWF, CRP, urate, PTH and serum lipids were higher in the uraemic patients compared to healthy
Pre-diabetes and arterial stiffness in uraemia
1223
Table 4. Univariate and multiple regression analysis for all uraemic subjects (n = 66)
Univariate regression
β
P
Partial r2
Multiple regression
Model R2 = 53%
Pulse wave velocity
Age (years)
MAP (mmHg)
Heart rate (bpm)
Gender (F/M)
Pre-diabetes (Y/N)
Dialysis (Y/N)
0.13
0.08
0.06
0.3
1.83
0.6
0.0001
0.005
0.15
0.71
0.03
0.66
0.34
0.17
0.05
0.00
0.11
0.00
<0.0001
0.004
0.71
0.54
0.11
0.71
Augmentation index
Age (years)
MAP (mmHg)
Heart rate (bpm)
Gender (F/M)
Height
Pre-diabetes (Y/N)
Dialysis (Y/N)
0.19
0.24
−0.41
11
−0.56
1.70
9.9
0.12
0.04
0.005
0.001
0.0004
0.60
0.09
0.04
0.07
0.13
0.18
0.19
0.005
0.05
Model R2 = 61%
0.0004
0.002
<0.0001
0.004
0.01
0.23
0.01
Data entered into the model included: Age (years), gender (F/M), MAP (mean arterial pressure, mmHg), Dialysis (Y/N), Pre-diabetes (Y/N) and insulin
resistance index (ISI). Bpm (beats per minute), Y = yes, N = no.
controls (Table 2), but no difference between pre-diabetic
and normoglycaemic uraemic patients was demonstrated.
Discussion
The main findings in this study were a high prevalence
(41%) of pre-diabetes among non-diabetic ESRD patients
on the waiting list for transplantation combined with findings, suggesting that increased aorta stiffness is more prevalent in pre-diabetic uraemic patients as compared to
uraemic patients with a normal glucose tolerance test.
The two uraemic groups were comparable regarding
age, time on dialysis and other key clinical and demographic parameters except for gender. Female sex was
more prevalent in the pre-diabetic group. This skewed distribution is, however, in keeping with the higher prevalence of females seen in population-based studies of
pre-diabetes [38], and the sex distribution in the uraemic
patients without pre-diabetes was similar to what is seen
in the total population of patients with ESRD in Denmark
[39]. Non-participation was mainly due to practical issues
such as long distance to the hospital and therefore does
not seem to bias our results. All tests had been made under standardized conditions. In particular, the influence of
dialysis status had been minimized because all investigations took place in the first days after hemodialysis, and in
patients treated with peritoneal dialysis, the peritoneal
cavity had been drained at 6 a.m. on the day of examination. The glucometabolic status was measured using the
golden standard of a glucose tolerance test with simultaneous measurements of p-insulin.
The reference range of aortic PWV in normal subjects
similar of age to our cohort is often reported to be below
6–8 m/s [18,40], and the presence of hypertension and
ESRD is associated with increased cardiovascular disease
and all-cause mortality [11,41]. A level of 9m/s as seen in
these uraemic patients with pre-diabetes may reflect an advanced stage of vascular disease in the arterial wall. It is
well known that the carotid-femoral PWV measured by
different methods is increased in patients with isolated
pre-diabetes and diabetes [42]. It is, however, a new observation that an increased carotid-femoral PWV in uraemic
patients may be related to pre-diabetes. Using univariate analysis, both AUCGlu120 min and presence of prediabetes was associated to PWV in uraemic patients, but the
multivariable analysis only disclosed age and blood pressure as independent determinants of PWV. When analysing
the three subgroups separately, we found that in the uraemic
patients with NGT and pre-diabetes as well as in healthy
controls, PWV was positively and independently associated
to age and MAP. In healthy controls, the 2-h glucose was
also positively (r = 0.78, P = 0.03) associated to PWV. Based
on these observations, it could be speculated if the higher
levels of blood pressure in uraemic patients may have
masked a possible effect of higher glucose levels on PWV.
This may find support in the observation that presence of
pre-diabetes and AUCGlu120 min was related to PWV in
the uraemic patients using the univariate analysis (P <
0.01). Furthermore, the association between pre-diabetes
and PWV reached a P level of 0.11 in the multivariate
analysis. We may therefore have overlooked significant associations due to the relatively small number studied.
The levels of AIX, SEVR and Tr were similar in uraemic patients with and without pre-diabetes and show that
other factors are responsible for the vascular pathophysiology in uraemia and the demonstrated differences in these
parameters in uraemic patients as compared to healthy
controls. Indeed, we found associations between increased
blood pressure, heart rate and age on both PWV and AIX,
which are in keeping with others [43], and indicate that
these parameters are important determinants of arterial
stiffness also in uraemic patients.
In addition, dialysis treatment was also independently associated to a higher AIX, indicating that the severity of ESRD
or the dialysis per se deteriorates arterial function [44].
The endothelial function in uraemia was significantly
reduced as indicated by an overall lowered FMD and in-
1224
creased von Willebrand factor antigen concentration in uraemic patients. However, a further reduction of endothelial
function in the uraemic patients with pre-diabetes was not
noted, suggesting that the uraemic state has a more pronounced impact on the endothelial dysfunction than presence of pre-diabetes.
PWV was positively associated to the diameter of the
brachial artery in the uraemic patients, but no other associations between measures of arterial stiffness and endothelial dysfunction were detected in this population.
Several measures of the metabolic syndrome such as
waist–hip ratio, plasma triglycerides and blood pressure
were increased in the uraemic patients, but no further increment was noted with presence of pre-diabetes. This is in accordance with a remarkable reduction in insulin sensitivity
in the uraemic patients with only a tendency to further reduction in the patients with pre-diabetes. The patients with
pre-diabetes had for the majority increased postprandial
glucose increment, and this could be due to an impaired incretin effect as it is known that patients with type 2 diabetes
and subjects with impaired glucose tolerance have reduced
postprandial GLP-1 secretion [45,46]. Uraemia per se could
have a deleterious effect on the GLP-1 secretion, which is
presently studied further in our laboratory. HbA1C was
not significantly elevated in our group of patients with
pre-diabetes. This might reflect that only a small deterioration in the glucose metabolism was seen or may be due to the
fact that a certain level of glucose is often reflected by a lower HbA1c level in uraemic patients than in healthy controls.
Finally, height was inversely associated to AIX as previously shown [47].
Based on our discussion above and the participation rate
of 48% in this population of chronic diseased patients, it is
likely that our findings can be generalized to all uraemic
patients accepted for kidney transplantation. Consequently,
the findings of our study suggest that increased PWV noted
in new onset diabetes after kidney transplantation [48,49]
may already, at least in part, be present before the transplantation and a result of pre-diabetes or insulin resistance in the
uraemic patients awaiting a kidney transplant.
In conclusion, a high prevalence of pre-diabetes, impaired insulin resistance, increased arterial stiffness of aorta
as well as impaired augmentation index and vasodilatation
was demonstrated in uraemic patients prior to kidney transplantation. Increased arterial stiffness of aorta and augmentation index were independently associated with age and
blood pressure. This supports the implementation and continuation of interventions aiming to reduce the cardiovascular load of risk factors including new onset diabetes mellitus
prior to kidney transplantation in uraemic patients.
Acknowledgements. We thank the laboratory technicians Annette Vinding, Helle Christensen and Andreas Haltorp for their skilful contribution
to the data collection and to Karl Bang Christensen, Department of Biostatistics, University of Copenhagen, for statistical advice. The study was
supported by an unrestricted grant from the Danish Kidney Foundation,
A.P. Møller Foundation for the Advancement of Medical Science, and the
Helen Bjørnow Foundation. M Hornum was supported by a fellowship
from Rigshospitalet, Denmark.
Conflicts of interest and disclosure statement. The results presented in this
paper have not been published previously in whole or part, except in abstract format. We have no conflicts of interest to disclose.
M. Hornum et al.
References
1. Cosio FG, Kudva Y, van d V et al. New onset hyperglycemia and
diabetes are associated with increased cardiovascular risk after kidney
transplantation. Kidney Int 2005; 67: 2415–2421
2. Hjelmesaeth J, Hartmann A, Leivestad T et al. The impact of earlydiagnosed new-onset post-transplantation diabetes mellitus on survival and major cardiac events. Kidney Int 2006; 69: 588–595
3. Shoji T, Emoto M, Shinohara K et al. Diabetes mellitus, aortic stiffness, and cardiovascular mortality in end-stage renal disease. J Am
Soc Nephrol 2001; 12: 2117–2124
4. Eldin WS, Ragheb A, Klassen J et al. Evidence for increased risk of
prediabetes in the uremic patient. Nephron Clin Pract 2008; 108:
c47–c55
5. DeFronzo RA, Smith D, Alvestrand A. Insulin action in uremia. Kidney Int Suppl 1983; 16: S102–S114
6. Eidemak I, Feldt-Rasmussen B, Kanstrup IL et al. Insulin resistance
and hyperinsulinaemia in mild to moderate progressive chronic renal
failure and its association with aerobic work capacity. Diabetologia
1995; 38: 565–572
7. Hager SR. Insulin resistance of uremia. Am J Kidney Dis 1989; 14:
272–276
8. Schmitz O. Peripheral and hepatic resistance to insulin and hepatic
resistance to glucagon in uraemic subjects. Studies at physiologic
and supraphysiologic hormone levels. Acta Endocrinol (Copenh)
1988; 118: 125–134
9. O'Rourke MF, Gallagher DE. Pulse wave analysis. J Hypertens Suppl
1996; 14: S147–S157
10. Schram MT, Henry RM, van Dijk RA et al. Increased central artery
stiffness in impaired glucose metabolism and type 2 diabetes: the
Hoorn Study. Hypertension 2004; 43: 176–181
11. Blacher J, Guerin AP, Pannier B et al. Impact of aortic stiffness on
survival in end-stage renal disease. Circulation 1999; 99: 2434–2439
12. Groothoff JW, Gruppen MP, Offringa M et al. Increased arterial stiffness in young adults with end-stage renal disease since childhood. J
Am Soc Nephrol 2002; 13: 2953–2961
13. London GM, Marchais SJ, Guerin AP. Arterial stiffness and function in
end-stage renal disease. Adv Chronic Kidney Dis 2004; 11: 202–209
14. Savage T, Clarke AL, Giles M et al. Calcified plaque is common in
the carotid and femoral arteries of dialysis patients without clinical
vascular disease. Nephrol Dial Transplant 1998; 13: 2004–2012
15. Airaksinen KE, Salmela PI, Linnaluoto MK et al. Diminished arterial elasticity in diabetes: association with fluorescent advanced
glycosylation end products in collagen. Cardiovasc Res 1993; 27:
942–945
16. Oxlund H, Rasmussen LM, Andreassen TT et al. Increased aortic
stiffness in patients with type 1 (insulin-dependent) diabetes mellitus.
Diabetologia 1989; 32: 748–752
17. Taniwaki H, Kawagishi T, Emoto M et al. Correlation between the
intima-media thickness of the carotid artery and aortic pulse-wave
velocity in patients with type 2 diabetes. Vessel wall properties in
type 2 diabetes. Diabetes Care 1999; 22: 1851–1857
18. Rahman S, Ismail AA, Ismail SB et al. Early manifestation of macrovasculopathy in newly diagnosed never treated type II diabetic patients with no traditional CVD risk factors. Diabetes Res Clin
Pract 2008; 80: 253–258
19. Cockcroft JR, Wilkinson IB, Evans M et al. Pulse pressure predicts
cardiovascular risk in patients with type 2 diabetes mellitus. Am J
Hypertens 2005; 18: 1463–1467
20. Schram MT, Kostense PJ, van Dijk RA et al. Diabetes, pulse pressure
and cardiovascular mortality: the Hoorn Study. J Hypertens 2002; 20:
1743–1751
21. Stehouwer CD, Henry RM, Ferreira I. Arterial stiffness in diabetes
and the metabolic syndrome: a pathway to cardiovascular disease.
Diabetologia 2008; 51: 527–539
22. Oflaz H, Pusuroglu H, Genchallac H et al. Endothelial function is
more impaired in hemodialysis patients than renal transplant recipients. Clin Transplant 2003; 17: 528–533
23. Henry RM, Ferreira I, Kostense PJ et al. Type 2 diabetes is associated
with impaired endothelium-dependent, flow-mediated dilation, but
Outcome of AF in ESRD patients
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
impaired glucose metabolism is not; the Hoorn Study. Atherosclerosis
2004; 174: 49–56
Celermajer DS, Sorensen KE, Gooch VM et al. Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. The Lancet 1992; 340: 1111–1115
Diagnosis and classification of diabetes mellitus. Diabetes Care
2007; 30 Suppl 1: S42–S47
Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from
oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 1999; 22: 1462–1470
Stumvoll M, Mitrakou A, Pimenta W et al. Use of the oral glucose
tolerance test to assess insulin release and insulin sensitivity. Diabetes
Care 2000; 23: 295–301
Stumvoll M, Van Haeften T, Fritsche A et al. Oral glucose tolerance
test indexes for insulin sensitivity and secretion based on various
availabilities of sampling times. Diabetes Care 2001; 24: 796–797
Buckberg GD, Fixler DE, Archie JP et al. Experimental subendocardial ischemia in dogs with normal coronary arteries. Circ Res 1972;
30: 67–81
Ferro G, Duilio C, Spinelli L et al. Relation between diastolic perfusion time and coronary artery stenosis during stress-induced myocardial ischemia. Circulation 1995; 92: 342–347
Savage MT, Ferro CJ, Pinder SJ et al. Reproducibility of derived central arterial waveforms in patients with chronic renal failure. Clin Sci
(Lond) 2002; 103: 59–65
Frimodt-Moller M, Nielsen AH, Kamper AL et al. Reproducibility of
pulse-wave analysis and pulse-wave velocity determination in chronic
kidney disease. Nephrol Dial Transplant 2008; 23: 594–600
Sidhu JS, Newey VR, Nassiri DK et al. A rapid and reproducible on
line automated technique to determine endothelial function. Heart
2002; 88: 289–292
Clausen P, Feldt-Rasmussen B, Iversen J et al. Flow-associated dilatory capacity of the brachial artery is intact in early autosomal dominant polycystic kidney disease. Am J Nephrol 2006; 26: 335–339
Anderson TJ, Uehata A, Gerhard MD et al. Close relation of endothelial function in the human coronary and peripheral circulations. J
Am Coll Cardiol 1995; 26: 1235–1241
Clausen P, Jensen JS, Jensen G et al. Elevated urinary albumin excretion is associated with impaired arterial dilatory capacity in clinically
healthy subjects. Circulation 2001; 103: 1869–1874
1225
37. Sorensen KE, Celermajer DS, Spiegelhalter DJ et al. Non-invasive
measurement of human endothelium dependent arterial responses:
accuracy and reproducibility. Br Heart J 1995; 74: 247–253
38. Balkau B, Hu G, Qiao Q et al. Prediction of the risk of cardiovascular
mortality using a score that includes glucose as a risk factor. The DECODE Study. Diabetologia 2004; 47: 2118–2128
39. Sorensen VR, Mathiesen ER, Heaf J et al. Improved survival rate in
patients with diabetes and end-stage renal disease in Denmark. Diabetologia 2007; 50: 922–929
40. Frimodt-Moller M, Nielsen AH, Kamper AL et al. Pulse-wave morphology and pulse-wave velocity in healthy human volunteers: examination conditions. Scand J Clin Lab Invest 2006; 66: 385–394
41. Laurent S, Boutouyrie P, Asmar R et al. Aortic stiffness is an independent predictor of all-cause and cardiovascular mortality in hypertensive patients. Hypertension 2001; 37: 1236–1241
42. Giannattasio C, Failla M, Capra A et al. Increased arterial stiffness in
normoglycemic normotensive offspring of type 2 diabetic parents.
Hypertension 2008; 51: 182–187
43. Schiffrin EL. Vascular stiffening and arterial compliance. Implications for systolic blood pressure. Am J Hypertens 2004; 17: 39S–48S
44. London GM, Drueke TB. Atherosclerosis and arteriosclerosis in
chronic renal failure. Kidney Int 1997; 51: 1678–1695
45. Toft-Nielsen MB, Damholt MB, Madsbad S et al. Determinants of
the impaired secretion of glucagon-like peptide-1 in type 2 diabetic
patients. J Clin Endocrinol Metab 2001; 86: 3717–3723
46. Vilsboll T, Krarup T, Deacon CF et al. Reduced postprandial concentrations of intact biologically active glucagon-like peptide 1 in type 2
diabetic patients. Diabetes 2001; 50: 609–613
47. Marchais SJ, Guerin AP, Pannier BM et al. Wave reflections and cardiac hypertrophy in chronic uremia. Influence of body size. Hypertension 1993; 22: 876–883
48. Kato K, Matsuhisa M, Ichimaru N et al. The impact of new-onset
diabetes on arterial stiffness after renal transplantation. Endocr J
2008; 55: 677–683
49. Saez AO, Kos M, Witzke O et al. Effect of new-onset diabetes mellitus on arterial stiffness in renal transplantation. Transplant Int 2008;
21: 930–935
Received for publication: 20.1.09; Accepted in revised form: 25.9.09
Nephrol Dial Transplant (2010) 25: 1225–1230
doi: 10.1093/ndt/gfp589
Advance Access publication 18 November 2009
Outcome of atrial fibrillation among patients with end-stage
renal disease
Che-Yi Chou, Huey-Liang Kuo, Shu-Ming Wang, Jiung-Hsiun Liu, Hsin-Hung Lin,
Yao-Lung Liu and Chiu-Ching Huang
Kidney Institute and Division of Nephrology, China Medical University Hospital, Taichung, Taiwan
Correspondence and offprint requests to: Yao-Lung Liu; E-mail: [email protected]
Abstract
Background. End-stage renal disease (ESRD) patients
are more at risk for atrial fibrillation (AF) than the general population. However, the prognosis in ESRD patients
with paroxysmal AF (PaAF), permanent AF (PAF) and
paroxysmal AF transformed to permanent AF (TAF) is
unknown.
Methods. In this retrospective longitudinal study, all
ESRD patients with PaAF, PAF and TAF between January
2001 and December 2007 were reviewed. The develop-
© The Author 2009. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.
For Permissions, please e-mail: [email protected]