Content uploaded by Satu Vehkavaara
Author content
All content in this area was uploaded by Satu Vehkavaara on Nov 16, 2015
Content may be subject to copyright.
Department of Medicine
Division of Diabetes
University of Helsinki
Helsinki, Finland
IDENTIFICATION AND TREATMENT OF
ENDOTHELIAL DYSFUNCTION AND
CARDIOVASCULAR RISK MARKERS IN
DISORDERS OF GLUCOSE METABOLISM AND IN
POSTMENOPAUSAL WOMEN
Satu Vehkavaara
ACADEMIC DISSERTATION
To be presented with the permission of the Medical Faculty of the University of Helsinki,
for public examination in auditorium 2, Meilahti Hospital, Haartmaninkatu 4, on June 8th,
2001, at 12 noon.
Helsinki 2001
Supervisor
Professor Hannele Yki-Järvinen
Department of Medicine
University of Helsinki
Helsinki, Finland
Reviewers
Professor John Cockcroft
Department of Cardiology
University of Wales College of Medicine
Cardiff, UK
and
Professor Olavi Ylikorkala
Department of Obstetrics and Gynecology
University of Helsinki
Helsinki, Finland
Official opponent
Professor Matti Tikkanen
Department of Medicine
University of Helsinki
Helsinki, Finland
ISBN 952-91-3415-0 (nid.)
ISBN 951-45-9960-8 (verkkojulkaisu, pdf)
http://ethesis.helsinki.fi
Yliopistopaino
Helsinki 2001
To Mikko
Abstract
Background and aims. Endothelial dysfunction precedes atherosclerosis. The relationship between novel
markers of cardiovascular risk such as impaired fasting glucose and the impact of treatment with common
therapeutic agents such as insulin and estradiol on endothelial function is unknown. The present studies were
undertaken to determine, whether i) impaired fasting glucose is associated with endothelial dysfunction; ii)
insulin therapy changes in vivo endothelial function in patients with type 2 diabetes, iii-v) estrogen
replacement therapy improves endothelial function, insulin action on glucose metabolism, peripheral blood
flow or arterial stiffness or markers of coagulation, fibrinolysis, inflammation and serum lipids and lipoproteins
and if so whether the improvement is dependent on the route of estradiol administration.
Subjects and methods. Endothelial function (forearm blood flow responses to intrabrachially infused
endothelium-independent and endothelium-dependent vasodilators) was measured in 17 subjects with
impaired fasting glucose and 12 subjects with normal fasting glucose concentrations (study I), and in 18 type
2 diabetic patients before and 6 months after bedtime insulin combination therapy and in 27 normal subjects
(study II). In studies III-V, 27 healthy postmenopausal women were randomized to receive either oral (n=9) or
transdermal (n=11) estradiol or placebo (n=7) for 12 weeks (studies III-V). Endothelial function was measured
after 0, 2, and 12 weeks of treatment. A euglycemic hyperinsulinemic clamp combined with measurement of
forearm blood flow, the augmentation index and peripheral vascular resistance was performed at baseline
and after 12 weeks of treatment. Markers of coagulation, fibrinolysis and inflammation, serum lipid,
lipoprotein, and apolipoprotein concentrations were measured and LDL particle size was quantitated in blood
samples taken at 0, 2 and 12 weeks of estrogen replacement therapy.
Results. Impaired endothelium-dependent vasodilatation was associated with altered glucose
homeostasis in subjects with impaired fasting glucose and in patients with type 2 diabetes. Insulin therapy
normalized both endothelium-dependent and -independent vasodilatation in patients with type 2 diabetes.
Oral but not transdermal estradiol improved endothelium-dependent and -independent vasodilatation in
forearm resistance vessels, increased markers of fibrinolytic activity and changed markers of coagulation
towards hypercoagulability, decreased soluble E-selectin concentrations and induced antiatherogenic
changes in serum lipids and lipoproteins, and increased serum C-reactive protein concentrations. Estradiol
increased peripheral blood flow, but had no effect on arterial stiffness or any action of insulin.
Conclusions. The finding of endothelial dysfunction in subjects with impaired fasting glucose supports the
idea that even mild abnormalities in glucose homeostasis may be markers of increased risk of cardiovascular
disease. Insulin therapy in type 2 diabetes has beneficial effects on endothelial function. The route of
estradiol administration is a critically important determinant of effects of estradiol on markers of
cardiovascular risk. Oral estradiol has multiple beneficial effects, but also possibly harmful effects.
Transdermal estradiol relieves postmenopausal symptoms but is neutral with respect to effects on endothelial
function, lipids and lipoproteins and markers of coagulation, fibrinolysis and inflammation.
CONTENTS
LIST OF ORIGINAL PUBLICATIONS ...................................................................................................... 7
ABBREVIATIONS ..................................................................................................................................... 8
1. INTRODUCTION ................................................................................................................................... 9
2. REVIEW OF THE LITERATURE .......................................................................................................... 10
2.1. The endothelial injury hypothesis .................................................................................................. 10
2.2. Assessment of vascular function in vivo: Methods and significance ............................................ 11
Resistance arteries
Physiological function
Endothelium-dependent and -independent vasodilatation
Cardiovascular risk factors and endothelial function
Lipids
Smoking
Hypertension
Cardiovascular disease and endothelial function
Large arteries
Physiological function
Methods to determine arterial stiffness
Cardiovascular risk factors and arterial stiffness
Cardiovascular disease and arterial stiffness
Vascular effects of insulin
Insulin-induced vasodilatation and changes in large artery stiffness
2.3. Impaired fasting glucose ............................................................................................................... 17
Definitions
Risk of cardiovascular disease
Markers of cardiovascular disease
2.4. Type 2 diabetes .............................................................................................................................. 18
Risk of cardiovascular disease
Endothelial function
Treatment of endothelial function
2.5. Hormone replacement therapy and cardiovascular disease .......................................................... 21
Epidemiology
Intervention studies
Effect of hormone replacement therapy on classic cardiovascular risk factors and novel
markers of cardiovascular risk
Lipids and lipoproteins
Endothelial function
Arterial stiffness
Antioxidants
Insulin resistance
Hormone replacement therapy and insulin sensitivity
Markers of inflammation, coagulation and fibrinolysis
Inflammation
Coagulation
Fibrinolysis
3. AIMS OF THE STUDY ........................................................................................................................... 28
4. SUBJECTS AND STUDY DESIGNS ..................................................................................................... 29
5. METHODS ............................................................................................................................................. 33
5.1. Vascular function ........................................................................................................................... 33
Endothelial function (Studies I-III)
Arterial stiffness (Study IV)
5.2. Insulin action .................................................................................................................................. 33
Glucose metabolism (Study IV)
Resistance arteries (Study IV)
Large arteries (Study IV)
5.3. Cardiovascular risk markers .......................................................................................................... 34
Markers of coagulation and fibrinolysis (Study V)
Marker of inflammation (Study V)
Marker of endothelial activation (Study V)
Serum lipids and lipoproteins (Study V)
Quantitation of LDL particle size (Study V)
5.4. Other measurements ..................................................................................................................... 35
Hormone concentrations (Studies I-V)
Measurement of total radical trapping capacity and water-soluble antioxidants (Study I)
Metabolic and physical parameters (Studies I-V)
5.5. Statistical analyses ........................................................................................................................ 36
6. RESULTS .............................................................................................................................................. 37
6.1. Endothelial function in subjects with impaired fasting glucose .......................................................37
6.2. Effect of insulin therapy on endothelial function ............................................................................. 38
6.3. Effect of oral and transdermal estrogen replacement therapy on endothelial function .................. 39
6.4. Effect of oral and transdermal estrogen replacement therapy on insulin sensitivity of glucose ..... 42
metabolism and preresistance and resistance vessel function in healthy postmenopausal
women
6.5. Effects of oral and transdermal estrogen replacement therapy on markers of coagulation, ..........45
fibrinolysis, inflammation and serum lipids and lipoproteins in healthy postmenopausal women
7. DISCUSSION ......................................................................................................................................... 48
7.1. Endothelial dysfunction and altered glucose homeostasis ............................................................ 48
Impaired fasting glucose
Type 2 diabetes
Treatment of endothelial dysfunction in patients with type 2 diabetes
7.2. Effects of oral and transdermal estradiol on markers of cardiovascular risk ................................. 50
Potentially beneficial effects
Potentially harmful effects
Cardiovascular risk markers not altered by estrogen replacement therapy
7.3. Concluding remarks ....................................................................................................................... 52
8. SUMMARY ............................................................................................................................................ 54
9. ACKNOWLEDGEMENTS ..................................................................................................................... 55
10. REFERENCES .................................................................................................................................... 56
ORIGINAL PUBLICATIONS
7
List of original publications
This thesis is based on the following publications, which are referred to in the text by their roman numerals:
I Vehkavaara S, Seppälä-Lindroos A, Westerbacka J, Groop P-H, Yki- Järvinen H: In vivo
endothelial dysfunction characterizes patients with impaired fasting glucose. Diabetes Care
22:2055-2060, 1999.
II Vehkavaara S, Mäkimattila S, Schlenzka A, Vakkilainen J, Westerbacka J, Yki-Järvinen H: Insulin
therapy improves endothelial function in type 2 diabetes. Arterioscler Thromb Vasc Biol 20:545-
550, 2000.
III Vehkavaara S, Hakala-Ala-Pietilä T, Virkamäki A, Bergholm R, Ehnholm C, Hovatta, O, Taskinen
M-R, Yki-Järvinen H: Differential effects of oral and transdermal estrogen replacement therapy on
endothelial function in postmenopausal women. Circulation 102:2687-2693, 2000.
IV Vehkavaara S, Westerbacka J, Hakala-Ala-Pietilä T, Virkamäki A, Hovatta O, Yki-Järvinen H:
Effect of estrogen replacement therapy on insulin sensitivity of glucose metabolism and pre-
resistance and resistance vessel function in healthy postmenopausal women. J Clin Endocrinol
Metab 85:4663-4670, 2000.
V Vehkavaara S, Silveira A, Hakala-Ala-Pietilä T, Virkamäki A, Hovatta O, Hamsten A, Taskinen M-
R, Yki-Järvinen H: Effects of oral and transdermal estrogen replacement therapy on markers of
coagulation, fibrinolysis, inflammation and serum lipids and lipoproteins in postmenopausal
women. Thromb Haemost 85:619-625, 2001.
The original publications are reproduced with permission of the copyright holders.
8
Abbreviations
ADMA asymmetric dimethylarginine
ACE angiotensin-converting enzyme
ACh acetylcholine
AT1angiotensin type 1
Apo apolipoprotein
BH4tetrahydrobiopterin
CEE conjugated equine estrogens
CHD coronary heart disease
CI confidence interval
CRP C-reactive protein
CV cardiovascular
EDHF endothelium-derived hyperpolarizing factor
ER estrogen receptor
ERT estrogen replacement therapy
FPG fasting plasma glucose
FFA free fatty acids
FSH follicle stimulating hormone
HbA1c glycosylated hemoglobin
HDL high density lipoprotein
HRT hormone replacement therapy
HERS Heart and Estrogen/progestin Replacement Study
IFG impaired fasting glucose
IGT impaired glucose tolerance
ICAM-1 intercellular adhesion molecule-1
LDL low density lipoprotein
L-NMMA NG-monomethyl-L-arginine
Lp lipoprotein
MPA medroxyprogesterone acetate
NO nitric oxide
NOS nitric oxide synthase
02-superoxide anion
OGTT oral glucose tolerance test
PAI-1 plasminogen activator inhibitor-1
PAP plasmin-antiplasmin complex
tPA tissue-type plasminogen activator
PEPI Postmenopausal Estrogen/Progestin Interventions
RR relative risk
SEM standard error of mean
SHBG sex hormone binding globulin
SNP sodium nitroprusside
TRAP total radical trapping capacity
UKPDS UK Prospective Diabetes Study
VCAM-1 vascular cell adhesion molecule-1
VLDL very low density lipoprotein
9
1. Introduction
Atherosclerosis is the underlying cause of heart disease and stroke and accounts for approximately half of all
deaths in Western societies. Over the past decades, major advances have been made in the understanding
of the pathogenesis of atherosclerosis. Epidemiological studies have revealed several important
environmental and genetic risk factors associated with atherosclerosis. The series of events in the vessel wall
that occur during atherogenesis are well defined. It has also become clear that the endothelium is an active
endocrine and paracrine organ the structural and functional integrity of which is critical for normal vascular
function. Also, blood-derived inflammatory cells play an important role in the pathogenesis of atherosclerosis.
Measurement of serum low density lipoprotein (LDL) and high density lipoprotein (HDL) cholesterol and
triglyceride concentrations and blood pressure have long been advocated as a way of identifying individuals
at increased risk of cardiovascular (CV) disease. During recent years several other risk factors have
emerged. These include alterations in concentrations of hemostatic factors, inflammatory markers and
possibly increases in glucose concentrations within the non-diabetic range. While it is clear that classic risk
factors of atherosclerosis and CV disease such as hypercholesterolemia and smoking cause endothelial
dysfunction, which is thought to be an early functional abnormality predisposing to atherosclerotic vascular
disease, the relationship between novel risk markers and endothelial function has been sparsely studied.
Little is also known of the impact of treatment of various risk factors on endothelial function. As an example,
insulin is widely used to treat patients with type 2 diabetes but no study has hitherto examined effects of
insulin therapy on endothelial function. Such study would seem important since insulin like other treatments
such as hormone replacement therapy (HRT) have effects on multiple parameters such as lipids and
markers of coagulation and fibrinolysis. New in vivo methods for measuring endothelial function in humans
provide potential tools for identification of individuals who should receive lifestyle modification and drug
therapies to prevent CV disease. With use of such methods and new biochemical markers for the
atherosclerosis, such as C-reactive protein (CRP), identification of high-risk individuals and testing of new
therapies might also be possible.
The present studies were undertaken to determine whether a slight elevation of fasting glucose (impaired
fasting glucose, IFG) is associated with in vivo endothelial dysfunction. In patients with type 2 diabetes,
effects of insulin therapy on endothelial function was determined. In postmenopausal women, effects of oral
and transdermal estrogen replacement therapy (ERT) on in vivo endothelial function and other novel markers
of CV risk were studied.
10
2. Review of the literature
2.1. The endothelial injury hypothesis
Over 150 years ago, Virchow postulated that a triad of conditions are needed to predispose to thrombus
formation, that is, abnormalities in blood flow, blood constituents, and the vessel wall. A modern viewpoint of
this triad includes abnormalities of hemorheology and turbulence at bifurcations and stenotic regions,
abnormalities in platelets and the coagulation and fibrinolytic pathways, and, finally, abnormalities in the
endothelium 1. Large epidemiological studies have revealed number of well established risk markers of
endothelial injury (Table 1).
The endothelium, a monolayer of elongated cells that lines all blood vessels, was long considered to be a
semipermeable membrane that prevented the diffusion of macromolecules. We now know that the
endothelium is the largest autocrine, paracrine and endocrine organ of the human body. It covers
approximately 700 m2 and weighs 1.5 kg; and regulates vessel tone, platelet activation, monocyte adhesion,
thrombogenesis, inflammation, lipid metabolism, and vessel growth and remodeling (Table 2) 2-5. The first
step in the pathogenesis of atherosclerosis is thought to be endothelial dysfunction that results from injury by
various CV risk factors and leads to inflammation and vessel remodeling 6.
Endothelial dysfunction that results from injury leads to compensatory responses that alter normal
homeostatic properties of the endothelium. Different forms of injury increase the adhesiveness of the
endothelium with respect to leukocytes or platelets, as well as its permeability. The injury also induces the
endothelium to have procoagulant properties and to form vasoactive molecules, cytokines, and growth
factors. If the inflammatory response does not effectively neutralize or remove the offending agents, it can
continue indefinitely. In doing so, the inflammatory response stimulates migration and proliferation of smooth-
muscle cells to form an intermediate lesion. The accumulation of monocytes within the subendothelium
constitutes the first stage of fatty streak. Continued inflammation results in an increased number of
macrophages and lymphocytes within the lesion 6. If these responses continue unabated, they thicken the
artery wall, which impedes blood flow 3. (Figure 1)
Continued inflammation results in the recruitment of increased numbers of macrophages and lymphocytes to
the lesion. Activation of these cells leads to the release of hydrolytic enzymes, cytokines, chemokines, and
growth factors 7; 8, which induce further damage and focal necrosis 9. Thus, cycles of accumulation of
mononuclear cells, migration and proliferation of smooth-muscle cells, and formation of fibrous tissue lead to
enlargement and restructuring of the lesion, which becomes covered by a fibrous cap that overlies a core of
lipid and necrotic tissue - an advanced, complicated lesion 6. At some point the artery can no longer
compensate by dilatation; the lesion then intrudes into the lumen and impedes blood flow 3. Plaque rupture
and thrombosis are complications of advanced lesions and cause unstable coronary syndromes or
myocardial infarction 9.
Table 1. Risk markers of endothelial injury.
• advanced age • inactivity
• male gender • high-fat diet
• elevated and modified LDL • family history of coronary heart disease
• smoking • genetic factors
• diabetes • elevated plasma homocysteine concentrations
• FFA and triglyceride-rich lipoproteins • infections
• hypertension
Table 2. Key regulatory functions of the vascular endothelium
• Semipermeable membrane • Monocyte adhesion
• Vascular tone • Inflammation
• Platelet aggregation and adhesion • Vessel remodeling and growth
• Thrombosis and thrombolysis • Lipoprotein metabolism
11
Foam
cells
F
atty
streak Intermediate
lesion
A
theroma
F
ibrous
plaque Complicated
l
esion/rupture
A
therosclerosis timeline
E
ndothelial dysfunction
From first decade From third decade From fourth decade
Growth mainly by lipid accumulation
S
mooth muscle
a
nd collagen Thrombosis,
haematoma
Figure 1. This graph (modified from ref. 10) illustrates the natural course of plaque formation. Early lesions in
the form of isolated macrophage foam cells may occur in infancy. Lipid accumulation can then lead to a fatty
streak. Next, lipids accumulate in the extracellular space within the vessel wall. After age 30, an atheroma or
visible lipid core may develop. At this point, plaque growth is marked primarily by lipid accumulation. From the
age of 40 onwards, plaques become more fibrous, a process which is dependent on the growth of a matrix of
smooth muscle cells and collagen over the atheromatous core. Finally, if unstable, plaques may erode or
rupture. Once the contents of the plaque are exposed to blood, platelet activation and thrombosis occur.
2.2. Assessment of vascular function in vivo: Methods and significance
Resistance arteries
Physiological function
The endothelium releases several agents that affect vascular smooth muscle function. Endothelium-derived
substances that relax the underlying smooth muscle include nitric oxide (NO), endothelium-derived
hyperpolarizing factor (EDHF), and prostacyclin 2; 11. Contracting factors include endothelin-1, angiotensin II,
thromboxane A2 and prostaglandin H2. NO is synthesized from L-arginine by NOS 12. Several isoforms of
nitric oxide synthase (NOS) have been identified, but only inducible and endothelial NOS are expressed in
endothelial cells 13. Endothelial NOS is responsible for the arterial tone at rest, and can be stimulated by
several receptor-dependent agonists (acetylcholine (ACh), methacholine, carbachol, thrombin, bradykinin,
substance P, and muscarinic agonists) and physical stimuli like shear stress. Activation of endothelial NOS is
Ca2+-dependent 14, while inducible NOS is activated independent of Ca2+ during inflammation by cytokines 15.
NO relaxes smooth-muscle cells through binding to guanylate cyclase and by increasing intracellular
concentrations of cyclic guanosine monophosphate 16. Vasodilator prostacyclin is formed from arachidonic
acid in endothelial cell 11. The same system produces also contacting factors like thromboxane A2 and
prostaglandin H2 17; 18. (Figure 2)
Endothelium-dependent and -independent vasodilatation
Of all molecules produced by the endothelium, NO has received most attention. Of note, however, there are
several other circulating markers of endothelial function such as endothelin-1, soluble E-selectin, vascular cell
adhesion molecule-1 (VCAM), and urinary excreted prostacyclin metabolites. The use of these markers is
limited because of the difficulty to define to what extent they are secreted by endothelium. Endothelium-
dependent and -independent vasodilatation are most commonly determined by measuring the vasodilatory
responses to endothelium-dependent pharmacologic or physiologic stimuli. Endothelium-dependent agonists
12
Endothelin-
system Cyclooxygenase-
system L-Arginine-
system
Contraction Contraction Relaxation
EndotheliumSmooth muscle
ET-1 TXA2PGI2NO
M
ACh Shear
stress
cGMP
L-NMMA
ADMA
SNP
EDHF
Figure 2. Endothelium-dependent relaxing and contracting systems. Endothelium releases several smooth
muscle contracting and relaxing agents. This figure also shows the vasoactive mechanisms of acetylcholine
(ACh), sodium nitroprusside (SNP) and NG-monomethyl-L-arginine (L-NMMA). ADMA=asymmetric
dimethylarginine, cGMP=cyclic guanosine monophosphate, EDHF= endothelium-derived hyperpolarizing
factor, ET-1=endothelin-1, M=muscarinic receptor, NO=nitric oxide, PGI2=prostacyclin, TXA2=thromboxane
A2. II=competitive blockade of NO synthesis.
include ACh, carbachol, methacholine, serotonin, bradykinin, thrombin, and substance P. Comparison of
responses to endothelium-dependent vasodilators with those to an endothelium-independent vasodilators
(such as sodium nitroprusside (SNP) or glyceryltrinitrate) forms the basis of an endothelial function test (Fig.
3.) In this test, drugs are administered into the brachial artery and blood flows are recorded simultaneously in
both arms 19; 20 (Fig. 3.). Changes in blood flow provide a measure of endothelial function at the level of
resistance vessels. Co-infusion of NG-monomethyl-L-arginine (L-NMMA) with ACh allows quantification of the
NO-dependent component of ACh-stimulated blood flow 21. The % decrease in blood flow during infusion of
L-NMMA alone provides another measure of NO-dependent endothelial function 22.
Use of ACh or other endothelium-dependent agonists is complicated because of their rapid degradation and
because their vasodilatory effect is not exclusively mediated via NO. The metabolic instability of ACh may
result in differential responses to this drug arising from anatomical rather than functional differences.
Vasodilator responses to ACh have been shown a greater dependence on resting blood flow and on forearm
length than those to SNP 23. Correction for forearm length has been shown to abolish the difference in
endothelial function between men and women 23. The vasodilatory responses to ACh can only partly (20-
40%) be inhibited by L-NMMA, an inhibitor of NOS 21; 22. In contrast, methacholine appears to vasodilate
through pathways other than L-arginine/nitric oxide 21. ACh also releases an EDHF 24, which causes
vasorelaxation of smooth muscle.
13
1000
500
0
1000
500
0
0204060
TIME (seconds)
CONTROL ARM
(mV) EXPERIMENTAL ARM
(ACh 15 µg/min)
FO R EAR M
BLO O D FLO W
2 ml/dl·min
FO R EAR M
BLO O D FLO W
10 ml/dl·min
Venous return
occluded Venous occlusion
released
Figure 3. Forearm blood-flows during infusion of ACh in the experimental (upper panel) and control arm
(lower panel). Arterial inflow is determined by drawing a tangential line across the first few pulses following
inflation of the sphygmomanometer cuff. The slope of this line reflects the volume change per unit time.
mV=millivolts.
Clinically significant atherosclerotic changes do not develop in the brachial artery. However, in an autopsy
study atherosclerotic endothelial lesions occurred commonly in the human brachial artery 25, and their
severity correlated significantly with those in the carotid and coronary arteries 25. Impaired endothelium-
dependent vasodilation of forearm resistance vessels has been shown to correlate with impaired
endothelium-dependent vasodilation in coronary arteries 26; 27. Improved endothelium-dependent dilatation
after cholesterol-lowering therapy has been shown in both forearm 28-31 and coronary 32; 33 circulation. A good
correlation has been found between flow-mediated dilatation of the brachial and coronary (intra-coronary
infusion of acetylcholine) arteries 34; 35 suggesting that peripheral arteries can be used in studies assessing
the predisposition to atherosclerosis in patients with cardiac risk factors. As discussed below, recent human
studies have also documented that endothelial vasodilator dysfunction is an independent predictor of vascular
events 36; 37.
Increased blood flow (shear stress) stimulates the endothelium mechanically and increases NO production.
Ultrasound measurement of changes in brachial artery diameter following induction of hyperemia with a blood
pressure cuff distal to the artery is another widely used method to test the capacity of endothelium to produce
NO 38. The endothelium-independent flow responses are tested with the use of sublingual glyceryltrinitrate. In
contrast to the intra-arterial method, this method is highly operator-dependent 39; 40.
Cardiovascular risk factors and endothelial function
Lipids
Multiple studies have documented that elevated serum levels of total and LDL cholesterol are associated with
endothelial dysfunction in forearm resistance 19; 41; 42 and coronary 43; 44 vessels independent of the presence
of coronary artery disease. Cholesterol-lowering therapy with statins reduces CV events 45-47 and total
mortality rates 45. Although statistically significant reductions in lesion severity have been demonstrated, the
extent of improvement has been considered modest relative to the observed clinical benefits, suggesting that
alternative mechanisms are operative, including plaque stability and improved endothelial function 48. Several
studies 32; 44; 49; 50 examined the effect of cholesterol-lowering therapy on coronary endothelial function and
demonstrated reduced coronary artery constriction 32; 44; 49; 50 and increased coronary blood flow responses
14
during intracoronary acetylcholine infusion 32; 33. Improved endothelium-dependent dilatation after cholesterol-
lowering therapy has also been shown in the forearm circulation 28-31. Antihypertensive effect of statins may
also contribute to the documented health benefits of these drugs 51. In contrast to these positive data, a
recent study where patients with mild coronary artery disease and mildly elevated cholesterol levels were
treated with simvastatin for 6 months failed to document improvements in coronary endothelial function 52.
Also combination therapy with gemfibrozil and (if necessary) niacin and/or cholestyramine in subjects with
normal or modestly elevated LDL cholesterol and low levels of HDL cholesterol had no effect on endothelial
function 53. Antioxidant therapy with vitamin E for 8 weeks did not reverse endothelial dysfunction in patients
with mild hypercholesterolemia and coronary artery disease 54. The influence of sex on endothelial function
has been previously studied 55. Results suggested women to be protected against adverse effects of
hypercholesterolemia on endothelial function since responses to ACh were impaired only in
hypercholesterolemic men but not in hypercholesterolaemic women 55.
In vitro, modified (mostly oxidized) LDL impairs endothelial function more than native 56. Oxidized lipoprotein
a (Lp(a)) has been suggested to impair endothelial function more than does oxidized native LDL in rabbit
renal arteries 57. The susceptibility of LDL to oxidation correlates better with impairment in endothelial function
than serum cholesterol concentration 58; 59. HDL cholesterol counteracts inhibitory effects of LDL on
endothelium-mediated vasodilation 60 and a positive correlation between HDL cholesterol and ACh-induced
coronary vasoreactivity has been observed 61. In patients free from other cardiac risk factors, modest chronic
elevation of triglycerides does not significantly attenuate flow-mediated dilation in the brachial artery 62 or
vasodilator responses to ACh 63. Small LDL particle size is associated with impaired endothelial function
independent of other lipid and lipoprotein concentrations 64.
Acute and chronic administration of vitamin C reverses endothelial dysfunction in the brachial circulations of
patients with coronary artery disease suggesting that increased oxidative stress contributes to endothelial
dysfunction in patients with atherosclerosis 65; 66. However, other studies in humans have shown no
improvement in endothelial function in response to administration of antioxidant vitamins 67; 68 or superoxide
dismutase 69. Hypercholesterolemia is also associated with increases in the circulating asymmetric
dimethylarginine (ADMA) (an endogenous inhibitor of NOS), which competes for substrate availability with L-
arginine 70.
Smoking
Cigarette smoking is a well established risk factor for atherosclerotic vascular disease 71, both in coronary
and peripheral arteries 72. Because nicotine impairs endothelium-dependent dilatation in human vessels in
vivo 73 and cigarette smoke contains a large number of oxidants 74, it has been proposed that the adverse
effects of smoking may result from oxidative damage to vascular endothelium. Indeed, endothelial
dysfunction in brachial 75 and coronary 76 arteries and coronary microcirculation 77 has been demonstrated in
long-term smokers and even in passive smokers 78; 79. Intra-arterial infusion of vitamin C 80; 81 and a single
oral dose of vitamin C 82 improves endothelium-dependent responses in chronic smokers, but oral vitamin C
therapy has no beneficial long-term effects 82. Oral supplementation of vitamin E can attenuate transient
impairment of endothelial function after heavy smoking but cannot restore chronic endothelial dysfunction in
healthy male smokers 83. On the other hand, vitamin E seems to improve endothelium-dependent relaxation
in forearm resistance vessels of hypercholesterolemic smokers, who have increased levels of autoantibodies
against oxidized LDL 84.
Hypertension
Lowering the blood pressure of hypertensive individuals decreases the incidence of stroke, heart and renal
failure, and mortality 85. ACh-induced relaxation has been impaired in the forearm of hypertensive humans in
some 20 86; 87 but not all 88 studies. Regular aerobic exercise for 12 weeks improves the forearm blood flow
response to ACh in normotensive as well as hypertensive subjects 89. Treatment of patients with essential
hypertension with angiotensin type 1 (AT1) receptor antagonist losartan but not with atenolol for one year has
been proposed to reverse functional changes of resistance arteries 90. Angiotensin-converting enzyme (ACE)-
inhibitors also seem to restore endothelial function in patients with hypertension in most 91; 92 but not all
studies 93. Results with calcium antagonists have been variable 91; 94; 95 and β-blockers 91 and diuretic agents
91; 94 have had no effect. Whether the putative beneficial vascular-protective effects of certain
antihypertensive drugs will translate into improved outcome in hypertension beyond the effect of blood
pressure lowering itself, is unknown.
15
Cardiovascular disease and endothelial function
Numerous studies have shown that paradoxical vasoconstriction induced by ACh occurs early as well as late
in the course of coronary atherosclerosis 96-98. Endothelial vasodilator dysfunction has been observed in
patients with traditional coronary risk factors, even in the absence of evidence for atherosclerotic lesions 99;
100. This supports the hypothesis that endothelial function reflects the impact of multiple coronary risk factors
on vascular function 101. If so, then coronary endothelial vasodilator dysfunction should predict coronary
disease progression and cardiovascular events. If this were the case, the assessment of endothelial
vasodilator function could become a useful prognostic tool in patients with coronary artery disease.
The close associations between CV risk factors and endothelial function do not prove endothelial dysfunction
to be a risk predictor of CV disease. Recently, two groups simultaneously demonstrated that endothelial
dysfunction indeed is an independent predictor of vascular events after adjustment for traditional CV risk
factors 36; 37. In the first study 36, 157 patients with mildly diseased coronary arteries who had undergone
measurement of coronary vascular reactivity by intracoronary ultrasound in response to intracoronary ACh,
adenosine, and nitroglycerin at the time of diagnostic study were followed for an average of 28 months.
Severe endothelial dysfunction, in the absence of obstructive coronary artery disease, predicted future
cardiac events 36. In the second study 37, coronary blood flow responses to ACh, sympathetic activation by
cold pressor testing, shear stress induced by papaverine or adenosine, and blood flow responses to
nitroglycerin were measured in 147 patients. CV events were recorded over a median follow-up period of 7.7
years 37. Patients suffering from CV events (n=16) had significantly increased vasoconstrictor responses to
ACh and cold pressor testing, and significantly blunted vasodilator responses to shear stress and
nitroglycerin 37. When the vasomotor responses of the different tests to assess coronary endothelial
vasoreactivity were entered into the multivariate analyses, classic risk factors except for hypertension were
no longer significant independent predictors of a worse clinical outcome 37. Moreover, coronary endothelial
vasodilator dysfunction remained an independent predictor of disease progression, even after controlling for
angiographic evidence of coronary atherosclerosis 37. These data further support the concept that endothelial
dysfunction has prognostic significance and is a predictor of CV events 101.
Large arteries
Physiological function
Blood pressure includes both static and dynamic component 102. Mean arterial pressure is dependent on
cardiac output and peripheral vascular resistance, and represents the static component of blood pressure.
Pulse pressure is influenced by arterial stiffness, stroke volume, and left ventricular ejection rate and
represents the dynamic component of blood pressure 102. In the normal vasculature, the large arteries act as
a ”buffering” system that is dependent on vessel compliance. During systole, the stroke volume is ejected into
the arterial tree 102. Part of the forward moving pressure wave is reflected back from reflectance sites which
are mainly located in large arteries 102. With arterial stiffening, which occurs during aging and is observed in
individuals with various risk factors (vide infra), the reflected wave returns earlier and augments systolic blood
pressure 103-106.
Methods to determine arterial stiffness
Arterial compliance can be estimated by measuring pulse wave velocity, pulse pressure, pressure and
contour of the pulse wave either centrally or by generating the central pressure waveform using a peripherally
recorded pressure wave and a transfer function (pulse wave analysis). Ultrasound techniques measure
changes in diameter relative to changes in pressure. Pulse wave velocity is applicable to large arterial
segments only and ultrasound techniques are limited by the ability of the method to image accurately the
vessel walls under investigation, and thus are only applicable to large accessible arteries. The technique of
pulse wave analysis reflects stiffness of the entire vasculature 107. Reproducibility of this method is similar or
better than that of other methods used to assess stiffness 108-110. However, if pulse wave analysis is used to
assess stiffness, heart rate and ejection duration need to be constant or adjusted for 111.
Cardiovascular risk factors and arterial stiffness
Stiffening of arteries is a consequence of the normal aging process 103; 112. In large arteries, with aging there
is a thinning and fracturing of elastin and increased collagen deposition resulting in increased wall thickness;
these changes adversely affect compliance 113; 114. Stiffening of arteries is accompanied by an increase in
16
systolic and pulse pressure, and an increased risk of CV morbidity and mortality. Recently published report
from the Framingham Heart Study demonstrated that with advancing age there was a gradual shift drom
diastolic blood pressure to systolic blood pressure and eventually to pulse pressure as predictors of coronary
heart disease (CHD) risk 115. This finding that pulse pressure and systolic blood pressure dominate as
predictors of CHD risk in the group of over 60 years of age is consistent with large artery stiffness contributing
to CHD risk in older hypertensives 116. In the Baltimore Longitudinal Study of Aging, changes in systolic
pressure, the carotid pulse augmentation index (measure of arterial stiffness), and the aortic pulse wave
velocity were reported in 146 male and female volunteers 21 to 96 years of age 117. Systolic blood pressure
increased 14%, aortic pulse wave velocity increased 2.5-fold, and the augmentation index increased 5-fold
over the age range studied 117. The rise in systolic blood pressure was similar in sedentary and endurance
trained individuals despite the 5-fold increase in the carotid pulse augmentation index in the sedentary group
and a 2-fold increase in the endurance-trained group 117. The use of a Doppler ultrasound method has
demonstrated age and sex-related differences in diameter and compliance in the abdominal aorta and
suggested that degenerative changes appear later in females than in males 118; 119. A gender difference has
also been found in an other study 120. Arterial compliance is also decreased in patients with CHD 104,
hypertension 105, familial hypercholesterolemia 106, and in those with type 2 diabetes 121. In cross-sectional
studies, aerobically trained athletes have a higher arterial compliance than sedentary individuals 117; 122.
Smoking and insulin resistance may also diminish stiffness 120; 123; 124.
Cardiovascular disease and arterial stiffness
Increased vascular stiffness is not just a marker for atheromatous vascular disease but is also an important
CV risk factor 125. An increase in pulse pressure is associated with an increase in cardiac morbidity and
mortality 126. In hypertensive patients, aortic pulse wave velocity has recently been shown to strongly
associate with the presence and extent of atherosclerosis and to constitute a powerful marker and predictor
of CV risk 127. Pulse pressure has been shown to predict recurrent events after myocardial infarction 128 and
increased mortality 129 in patients with impaired left ventricular function, and coronary events in untreated
hypertensive male subjects 130. Increased pulse pressure has also been shown to be an independent
predictor of the incidence of CHD and overall mortality among elderly 131.
Vascular effects of insulin
Insulin-induced vasodilation and changes in large artery stiffness
In addition to its action on glucose metabolism, insulin has hemodynamic effects. The ability of insulin to
increase peripheral blood flow is dependent upon the duration and dose of insulin exposure 132. At
physiological insulin concentrations such as those prevailing during an intravenous insulin infusion at a rate of
1 mU/kg⋅min, however, insulin has no or only minor effects on peripheral blood flow 132; 133. Insulin-induced
vasodilation can be abolished with L-NMMA 134; 135 and ouabain 136. Until recently it was unknown whether
insulin affects function of arteries greater than those regulating peripheral vascular resistance, i.e. resistance
arteries. Westerbacka et al. 110 was the first to demonstrate that insulin acutely diminishes arterial stiffness
independent of changes in peripheral vascular resistance. Within an hour, insulin reduced wave reflection at
the level of aorta. Peripheral systolic and diastolic blood pressure, blood flow, and vascular resistance did not
change significantly until 2 to 3 hours after the start of the insulin infusion 110. These data suggest hierarchy in
the vascular effects of insulin and suggest that insulin increases the diameter of distensibility of arteries
before resistance vessels. Defects in these vascular actions of insulin have been suggested to provide a
novel mechanistic link between insulin resistance and systolic hypertension 137.
17
2.3. Impaired fasting glucose
Definitions
In 1997, an Expert Committee of the American Diabetes Association proposed modifying the diagnostic
criteria for diabetes, by lowering the fasting plasma glucose at which diabetes can be diagnosed from 7.8 to
7.0 mmol/l 138.The Expert Committee recognized an intermediate group of subjects whose glucose levels,
although not meeting criteria for diabetes, were too high to be considered altogether normal. This group was
defined as having fasting plasma glucose (FPG) levels ≥ 6.1 mmol/l but < 7.0 mmol/l or 2-h values in the oral
glucose tolerance test (OGTT) of ≥ 7.8 mmol/l but < 11.1 mmol/l. (Figure 4)
Risk of cardiovascular disease
CV complications are often present already at the diagnosis of type 2 diabetes 139. Subjects with impaired
glucose tolerance (IGT) have an approximately twofold increase in the risk of macrovascular disease 139.
Epidemiological evidence, such as those generated in the UK Prospective Diabetes study (UKPDS), have
shown that the risk of CV disease increases linearly with increasing glycemia 140. In the Rancho Bernardo
Study 141, an increase of FPG from 5 to 7 mmol/l was associated with a doubling of CHD mortality in men and
a tripling in women. The Paris Prospective Study reported that the risk of developing diabetes over 3 years
was greater among middle-aged men with a FPG > 6.1 mmol/l than it was in those with a lower FPG 142.
Within the same cohort, it has also been reported that CHD mortality is elevated among people with a FPG in
the range from 5.8 to 6.9 mmol/l 143. Coutinho et al. 144 have recently published a meta-regression analysis of
20 studies including 95783 non-diabetic individuals who had 3707 cardiovascular
Stages
Types
Normoglycemia Hyperglycemia
Type 1*
Type 2
Other Specific Types **
Gestational Diabetes **
Normal Glucose Regulation Diabetes Mellitus
Not insulin
requiring Insulin requiring
for control Insulin requiring
for survival
IFG IGT
Fasting plasma glucose, mmol/l
2-hr post glucose load, mmol/l
< 6.1
< 7.8
6.1-6.9
< 7.8
< 7.0
7.8-11.0
≥ 7.0
> 11.0
Figure 4. Disorders of glycemia:etiologic types and stages (modified from ref.138). * Even after presenting in
ketoacidosis, these patients can briefly return to normoglycemia without requiring continuous therapy (i.e.,
”honeymoon” remission). ** In rare instances, patients in these categories (e.g., type 1 diabetes presenting in
pregnancy) may require insulin for survival.
18
events and were followed for 12.4 years. Compared with a glucose level of 4.2 mmol/l, a fasting glucose level
of 6.1 mmol/l and 2-hour glucose level of 7.8 mmol/l were associated with a relative CV event risk of 1.33
(95% confidence interval (CI) 1.06-1.67) and 1.58 (95% CI 1.19-2.10), respectively. Recently published
results from Norfolk cohort of European Prospective Investigaton of Cancer and Nutrition (EPIC-Norfolk)
showed glycosylated hemoglobin (HbA1C) concentrations to predict mortality continuously across the whole
population distribution in people without diabetes and at concentrations below those used to diagnose
diabetes 145.
Markers of cardiovascular disease
A population-based survey on the island of Mauritius with a follow-up of 5 years showed that at baseline,
blood pressure, lipids, and obesity increased in a linear fashion with increasing FPG, with no evidence of a
threshold effect 146. Intima-media thickness is increased in individuals with IGT 147 and correlates positively
with FPG concentrations 148. Yamasaki et al. 149 have shown that asymptomatic hyperglycemia is associated
with increased carotid artery intima-media thickness in non-diabetic subjects. In subjects with IFG, the
microvascular hyperemic response to local heating of the skin of the foot has been found to be blunted
compared with subjects with normal fasting glucose 150. Abnormalities in vascular reactivity as measured by
iontophoresis are present in individuals at risk of developing type 2 diabetes (subjects with IGF or with a
positive family history of type 2 diabetes in one or both parents) 151. In 60 men with various CV risk factors,
the fasting glucose concentration was the only variable that was correlated both with postischemic,
endothelium-dependent vasodilation and increased intima-media thickness 152. These data suggest that
impaired fasting glucose could be an early marker of vascular dysfunction. It is not known, however, whether
endothelial dysfunction characterizes subjects with IFG with or without CV risk factors as no studies have
measured endothelial function in subjects with IFG.
2.4. Type 2 diabetes
Risk of cardiovascular disease
CV disease is the most common complication and leading cause of death in type 2 diabetes 153; 154.
Epidemiological studies show that the risk of CV mortality is two to three times higher in men and three to
five times higher in women with diabetes than in non-diabetic subjects 155-160. The age-adjusted prevalence of
CHD in white diabetic adults is about 45%, compared with about 25% in non-diabetic individuals 161. CV
disease accounts for about 70% of all deaths in patients with diabetes 162. Diabetic patients without previous
myocardial infarction have as high a risk of myocardial infarction as non-diabetic patients with previous
myocardial infarction 163. Furthermore, diabetic patients with myocardial infarction have a much worse
prognosis than non-diabetic patients with myocardial infarction 164; 165.
Classic CV risk factors (elevated cholesterol concentration, smoking, hypertension) are important in type 2
diabetes, although they do not explain the excessive risk of CV disease. Although the LDL cholesterol
concentration is usually normal in type 2 diabetes, it is a strong predictor of the risk of CV events 155. Findings
in large lipid-lowering and antihypertensive trials suggest that lowering LDL cholesterol and blood pressure
reduces CV events in diabetic patients 46; 163; 166. In the Scandinavian Simvastatin Survival Study 45, lipid-
lowering therapy produced a greater reduction in the rate of coronary events in diabetic subjects than in non-
diabetic subjects (55% vs. 32%, respectively). In the Cholesterol and Recurrent Events study 46, there were
similar relative reductions in diabetic and non-diabetic subjects (27% vs. 25%, respectively) although the
absolute reduction was greater because of a higher event rate in diabetic than in non-diabetic participants.
The UKPDS showed that intensive blood pressure control clearly decreased the risk of both macro- and
microvascular events 167. Similarly, in participants with diabetes in the Systolic Hypertension in the Elderly
study, a decrease in systolic and diastolic pressures of 10 mmHg and 2 mm Hg, respectively, reduced the
risk of CV events by up to 34% 168. In the Heart Outcomes Prevention Evaluation study ACE inhibitor ramipril
significantly lowered the risk of major CV outcomes by 25-30% in a broad range of high-risk middle-aged and
elderly people with type 2 diabetes 169. This effect seemed partly independent of blood pressure lowering 169.
Several studies have indicated that hyperglycemia is an independent predictor of CV disease. The San
Antonio Heart Study demonstrated that hyperglycemia is a risk factor not only in Caucasians, but also in other
ethnic groups 170. The Wisconsin Epidemiologic Study of Diabetic Retinopathy assessed the significance of
glycemic control for micro- and macrovascular complications in 1370 subjects with late-onset diabetes during
10 years of follow-up 171. A 1% increase in HbA1c was associated with a 10% increase in CHD events. The
UKPDS, which included 3867 patients with newly-diagnosed type 2 diabetes aged 25-65 years, showed that
intensive blood glucose control with insulin or sulphonylureas retards the development of microvascular
19
complications 172. The incidence of myocardial infarction decreased by 16%, which was almost statistically
significant. Neither insulin nor sulphonylureas had adverse effects on cardiovascular outcome 172. In contrast
to these drugs, a substudy of the UKPDS suggested metformin to be cardioprotective in overweight patients
173. The UKPDS also evaluated the significance of all major CV risk factors for CHD by stepwise multivariate
Cox analysis 172. The most important risk factor for CHD was high LDL cholesterol, followed by low HDL
cholesterol and HbA1c.
Endothelial function
In patients with type 2 diabetes, an impaired vasodilator response to endothelium-dependent vasodilators,
such as ACh, has been a consistent finding 174-189. Endothelium-independent vasodilation has been either
impaired 177-179; 189-191 or normal 175; 176; 180; 181; 183-188; 192; 192. Table 3 summarizes results of in vivo endothelial
function tests in patients with type 2 diabetes. The key features of the studies, and the main results regarding
endothelial function and the method used are also shown. The studies are listed in chronological order and
according to method used.
Multiple causes could contribute to endothelial dysfunction in patients with type 2 diabetes compared with
non-diabetic subjects matched for traditional causes of endothelial dysfunction, such as age, cholesterol
concentrations and blood pressure. Such factors could include those known to be associated with increased
cardiovascular risk, such as chronic hyperglycemia 170; 171, hyperinsulinemia independent of insulin resistance
193, insulin resistance and its consequences 194 (hypertriglyceridemia, elevated blood pressure, increased
concentrations of small dense LDL particles and qualitative abnormality of LDL 195, low HDL cholesterol,
central obesity, increases in free fatty acids (FFA) concentrations, abnormal regulation of autonomic function
by insulin and insulin resistance of platelet function as well as abnormalities in coagulation parameters). Data
on possible causes of endothelial dysfunction are sparse in previous cross-sectional studies, but factors
found to be correlated with endothelial dysfunction include serum triglycerides (positively) 177, serum HDL
(inversely) 177, and LDL 175 cholesterol concentrations (positively), LDL particle size (positively) 178; 181, body
mass index (positively)175, and presence of peripheral sensory neuropathy (inversely)189.
Treatment of endothelial function
Regarding treatment of endothelial function in type 2 diabetes, most clinical studies to date have
concentrated on studying effects of antioxidant administration. Acute intra-arterial infusion of vitamin C was
found to augment endothelium-dependent vasodilation to methacholine by 50% 176, but there are no studies
addressing effects of chronic vitamin C therapy on endothelial function in type 2 diabetes. Oral treatment with
raxofelast, a new watersoluble vitamin-E-like antioxidant agent, for 1 week reduced oxidative stress and
improved blood flow responses to ACh in men with type 2 diabetes 184. Oral vitamin E supplementation (1600
IU daily for 8 weeks) has, however, been found not to improve endothelial dysfunction in uncomplicated type
2 diabetes 180. One way to prevent xanthine oxidase-generated free radicals is to use the xanthine oxidase
inhibitor allopurinol 196. Recently, data have shown that acute intra-arterial infusion of oxypurinol, the active
metabolite of allopurinol, improves endothelial function in hypercholesterolemic humans 197. Treatment with
allopurinol for 1 month has also been found to improve endothelial function in patients with type 2 diabetes
and mild hypertension 183. Intravenous infusion of L-arginine, a precursor of nitric oxide, did not improve
coronary dilation in diabetic patients, while deferoxamine, an ion chelator that prevents iron-catalyzed
generation of hydroxyl radicals, did so 198. Further studies in diabetic patients are required to establish
whether antioxidant therapy has long-term beneficial effects on endothelial function and whether it retards
development of CHD.
Tetrahydrobiopterin (BH4) is a cofactor of NOS, and may play a key role in the control of the calcium-
dependent production of NO and 02- in vivo 199. Its deficiency leads to uncoupling of the L-arginine-NO
pathway, resulting in increased formation of oxygen radicals by NOS and reduced NO production in vitro 200.
Oral administration of BH4 prevents endothelial dysfunction and vascular oxidative stress in the aortas of
insulin-resistant rats 201. Intra-arterial infusion of BH4 was also recently shown to improve endothelium-
dependent vasodilation in patients with type 2 diabetes 185 and in hypercholesterolemia 202. Single-dose of
orally administered saptopterin hydrochloride, an active analogue of BH4, restores endothelial function in
20
Table 3. Studies of in vivo endothelial function in patients with type 2 diabetes (DM2) compared to normal subjects (Cont).
Authors Subjects
(women/men) Glycemic control of
diabetes Duration of diabetes (years) Endothelium-dependent
agent and response vs Cont Endothelium-independent
agent and response vs Cont
Intra-arterial infusion and pletysmography method
McVeigh et al. (1992) DM2 (5/24)
Cont (5/16) HbA1C 9.7% (6.3-12.9%) 5 (1-16) ACh ↓
L-NMMA NS SNP ↓
Steinberg et al. (1996) DM2 (8)
Cont (13) -- MCh ↓SNP NS
Ting et al. (1996) DM2 (4/6)
Cont (5/5) HbA1C 7.9±0.7% 3 (1-7) MCh ↓SNP NS
Verapamil NS
Watts et al. (1996) and
O’Brien et al. (1997) DM2 (0/29)
Cont (0/18) HbA1C 7.5% (5.5.-10.8%) 4 (1-10) ACh ↓
L-NMMA NS SNP ↓
Williams et al. (1996) DM2 (7/14)
Cont (6/17) HbA1C 11±1% 4 (0.5-12) MCh ↓SNP ↓
Verapamil NS
Avogaro et al. (1997) DM2 (0/10)
Cont 6 HbA1C 8.7±0.6% 7±2ACh NS SNP NS
Gazis et al. (1999) DM2 (12/36)
Cont (11/10) HbA1C 6.9±1.4% 5±3 ACh ↓
Bk NS SNP NS
Mäkimattila et al. (1999) DM2 (0/30)
Cont (0/12) HbA1C 7.4±0.3% 4±1 ACh ↓
L-NMMA NS SNP NS
Cleland et al. (2000) DM2 (0/9)
Cont (0/9) - - L-NMMA NS norepinephrine NS
Steinberg et al. (2000) MCh:DM2 (7/8) Cont (5/44)
SNP:DM2 (5/3) Cont (0/18) -- MCh ↓SNP NS
Butler et al. (2000) DM2 + hypertension
(1/10) Cont (0/12) HbA1C 7.1±1.7% 1-20 (median 4) ACh ↓SNP NS
Chowienczyk et al.
(2000) DM2 (0/10)
Cont (0/10) HbA1C 8.1±2.4% -ACh ↓SNP NS
Heitzer et al. (2000) DM2 (7/16)
Cont (4/8) HbA1C 7.8±0.2% 5±1 ACh ↓
L-NMMA ↓SNP NS
Ultrasound method
Goodfellow et al. (1996) DM2 (6/6)
Cont (6/6) - 4 (1-7) FMD ↓GTN NS
Huvers et al. (1997) DM2 (5/13)
Cont (8/10) HbA1C 6.5% (6.1-7.8%) 4±2-GTN ↓
Enderle et al. (1998) DM2 (11/14)
Cont (11/14) HbA1C 9.1±2.4% 7±6 FMD ↓GTN NS
Iontophoresis method
Morris et al. (1995) DM2 (0/14)
Cont (0/14) HbA1C 6.5±0.2% 9±2 ACh ↓SNP ↓
Pitei et al. (1997) DM2 (3/4) DM2(PN) (3/5)
Cont (4/6) HbA1C 9.1±1.9% and
9.9±1.0% (PN) -DM2: ACh ↓
DM2 (PN): ACh ↓DM2: NS
DM2 (PN): SNP ↓
GTN=glyseryl trinitrate, Bk=bradykinin, PN=peripheral sensory neuropathy. ↓ = blunted response. Data are shown as mean±SEM.
21
smokers 203, but further studies are required to clarify the usefulness of BH4 treatment for the prevention of
endothelial dysfunction and the development of CV diseases in insulin-resistant states. ACE inhibitors as well
as angiotensin II blockade with losartan have improved endothelial function in type 2 diabetics in most 204-206,
but not all 207 studies. Fibrates are a widely used for lowering of lipids, exerting a variety of effects on lipid and
lipoprotein metabolism 208, particularly attenuation of postprandial lipemia 209 and reduction in triglyceride-rich
very low density lipoprotein (VLDL), the key abnormalities in type 2 diabetes 210. Ciprofibrate therapy, by
attenuating postprandial lipemia and modifying an atherogenic lipoprotein profile, caused significant
improvement in fasting and postprandial endothelial function and attenuated postprandial oxidative stress in
type 2 diabetes 211. Whether improvement of glycemic control with the use of oral medication, insulin or
combination therapy improves endothelial function has not been studied.
2.5. Hormone replacement therapy and cardiovascular disease
Epidemiology
Cardiovascular disease is the leading cause of death of American women 212. The weight of evidence from
case-control, cross-sectional, and prospective cohort studies has found the risk of CHD to be lower in
estrogen users than in non-users 213. Six hospital-based case-control studies found a summary relative risk
(RR) of 1.33 (95% CI 0.93-1.91), but the use of hospitalized patients as controls complicated interpretation of
these data 214. Nine population or community-based case-control studies indicated that HRT use is
associated with a relative risk reduction of 24% (RR 0.76, 95% CI 0.66-0.88) 214. Four cross-sectional
angiographic studies suggested an even greater risk reduction for CHD in users vs. non-users of HRT (RR
0.39, 95% CI 0.33-0.47) 214. In 16 prospective cohort studies, estrogen ever-users vs. never-users had a
pooled RR 0.70 (95% CI 0.63-0.77) for CHD 214. Combining all above study types, the pooled RR of ever-
users vs. never-users was 0.50 (95% CI 0.45-0.59) 213. Another meta-analysis also found a RR of 0.65 (95%
CI 0.59-0.71) for risk of CHD in ever-users vs. nonusers 215. The largest of the prospective studies, the
Nurses Health Study, in which 70 533 women have been followed for 20 years, found an adjusted RR for
coronary events of 0.54 (95% CI 0.44-0.67) in estrogen users vs. never-users 216. Furthermore, 0.3 mg of oral
conjugated equine estrogens (CEE) daily was associated with a reduction similar to that seen with the
standard dose of 0.625 mg 216. Also in Finland Sourander et al. 217 recently published large study in which
7944 women were followed from 1987 until 1995 and showed that current ERT was associated with a
reduction in sudden cardiac death and reduced mortality (RR 0.21, 95% CI 0.08-0.59). Taken together the
epidemiological evidence favoring use of HRT in postmenopausal women is overwhelmingly convincing.
Progestins are commonly added to protect against endometrial hyperplasia which would result from
unopposed use of estrogens. The addition of progestins to estrogen replacement therapy may have an
adverse effect on the serum lipid profile 218. Such data have raised the possibility that use of combined
therapy compared to estrogen-alone therapy in postmenopausal women may have smaller vascular benefits.
On the other hand, cross-sectional analysis of the Atherosclerosis in Communities study suggested that
women using estrogen with progestin had an even greater vascular benefit than those using estrogen alone
219. In the Nurses Health Study, CHD risk was similarly lower both in users of combined HRT (adjusted RR
0.39, 95% CI 0.19-0.78) and estrogen alone (adjusted RR 0.60, 95% CI 0.43-0.83) vs. never-users 220. In a
meta-analysis, the estimated RR of CHD with combined HRT was 0.65-0.80 215. In observational studies the
risk of myocardial infarction was comparably reduced with combined HRT as with estrogen alone 221; 222.
Intervention studies
Many observational studies have found lower rates of CHD in postmenopausal women who use estrogen as
compared to non-users 214; 215; 222. If this association is causal, estrogen therapy could prevent CHD in
postmenopausal women. It was therefore unexpected that the Heart and Estrogen/progestin study (HERS)
found no overall benefit of 4.1 years of treatment with CEE (0.625 mg) plus medroxyprogesterone acetate
(MPA) (2.5 mg) on the risk of nonfatal myocardial infarction and death from CHD among women with
established coronary atherosclerosis 223. Interpretation of these data are difficult since there was an early
increase and a late reduction in risk 223. Several factors have been proposed to explain these results such as
short duration of this trial, limited statistical power, impact of too high a dose of estrogen, multipharmacy,
possible negative effects of the progestin used and previously unrecognized or underemphasized
prothrombotic or proinflammatory effects of HRT 224; 225.
The Estrogen Replacement and Atherosclerosis trial was a randomized, double-blind, placebo-controlled
clinical trial that examined the effects of HRT on the progression of coronary atherosclerosis in women 226. A
total of 309 postmenopausal women who had angiographically verified CHD at baseline were randomly
22
assigned to receive unopposed estrogen (oral CEE 0.625 mg), estrogen plus MPA (2.5 mg), or placebo. A
mean of 3.2 years of ERT did not slow the progression of coronary atherosclerotic lesions in any study group
226. The lack of an treatment effect can be explained by the possibility that estrogen has proinflammatory
effects that offset its beneficial effects. Another explanation is that estrogen is more effective in preventing
the development of atherosclerosis than in slowing its progression. This study does not exclude the possibility
that estrogen therapy may still be beneficial in the primary prevention of CHD, but this has not yet been
verified. The Women’s Health Initiative, a large clinical trial including predominantly healthy women, will
provide important data about estrogen use for the primary prevention of heart disease 227.
Effect of hormone replacement therapy on classic cardiovascular risk factors and novel markers of
cardiovascular risk
Lipids and lipoproteins
After menopause, adverse lipid changes occur. These changes include increases in LDL and HDL3
cholesterol concentrations and decreases in HDL and HDL2 concentrations 228. Oral estrogen lowers LDL and
increases HDL cholesterol concentrations in postmenopausal women 218; 223; 229-233. Estrogen increases HDL2
much more than HDL3 cholesterol concentrations 229; 232; 234. In the Postmenopausal Estrogen/Progestin
Interventions (PEPI)-trial 218, in which 875 women were followed for 3 years, 0.625 mg/d of CEE resulted in a
9% rise in HDL cholesterol and a 10% fall in LDL cholesterol. These lipoprotein effects are seen only with oral
estrogen due to its first-pass hepatic effect 229. Estrogen enhances LDL catabolism 229 by increasing LDL
receptors on the cell surface 235. An important aspect of HDL metabolism is the fact that essentially all HDL
particles have apo A-I, but many lack apo A-II 236. HDL that contains apo A-II is called Lp A-I/A-II and appears
to be distinct in many ways from HDL without apo A-II, which is called lipoprotein (Lp) A-I. LpA-I may be a
better cholesterol acceptor in vitro 237 and appears to be more antiatherogenic than Lp A-I/A-II in cross-
sectional epidemiological studies 238. Oral ERT selectively raises Lp A-I levels apparently due to selective
elevation of the Lp A-I production rate 234; 239, and downregulates hepatic lipase activity 234; 239; 240, and may 241
or may not 239 reduce the fractional catabolic rate. Estradiol levels achieved during administration of 1 mg
estradiol were recently shown to positively correlate with increases in HDL cholesterol concentrations 242.
HDL is a more potent predictor of lower CHD risk in women than in men 243; 244. Estrogen’s HDL cholesterol
raising effect may thus lower the risk of CHD. Estrogen lowers LDL cholesterol, but its importance as a
predictor of primary CHD in women is unclear 244. In secondary prevention trials 45; 46, LDL cholesterol
lowering has similar, if not greater 46, benefits in women as in men.
Contrary to most previous studies 233; 245, all HRT groups in the PEPI-trial had significantly increased
triglyceride levels compared to non-users of HRT. The triglyceride increase did not differ between women
assigned to estrogen alone or with a progestin 218. In HERS mean triglyceride levels increased by 10% in the
hormone using group 223. Estrogen increases triglyceride levels by increasing triglyceride production rather
than impairing clearance, and these changes may not be atherogenic 246. ERT has been shown to increase
apolipoprotein (apo) AI and decrease apo B concentrations 242; 247. These effects appear greatest in
dyslipidemic women 247. Transdermal estradiol appears to have no 229; 233; 248; 249 or only minor 250-252 effects
on plasma lipids. Recently published results from the HERS suggest that Lp(a) is an independent risk factor
for recurrent CHD in postmenopausal women and that treatment with oral CEE and MPA lowers Lp(a) levels
253 as has also been previously shown 254. Oral HRT appeared to have a more favourable effect in women
with high initial Lp(a) levels than in women with low levels 253, a result also previously reported 251.
Transdermal estradiol may also decrease Lp(a) concentrations 251. Several recent studies strengthen
evidence that Lp(a) is a substantial risk factor for women younger than 65 years of age. Orth-Gomer et al. 255
found that median Lp(a) was 38% higher in patients hospitalized for an acute myocardial infarction or angina
pectoris compared with control subjects. In the prospective Framingham Study an elevated Lp(a)
concentration was an independent risk factor for cardiovascular disease for women (mean age, 45 years) 256.
It is unknown whether a reduction of Lp(a) diminishes the risk for cardiovascular disease in women. For
interested readers more detailed reviews of effects of HRT on lipids and lipoproteins are recommended 257;
258.
Endothelial function
Several studies indicate that tissues not classically defined as estrogen targets also express estrogen
receptor, including vascular smooth muscle cells 259; 260. Estrogen receptor has also been identified in
cultured human endothelial cells 261; 262 supporting the hypothesis that cardioprotective effects of estrogen are
mediated, at least in part, through a classic steroid hormone receptor mechanism. Estrogen receptor (ER)-α
has long been considered to be the unique target of estradiol, because it has been characterized in
23
reproductive tissues, and in vessels, particularly in the endothelial cells 261; 263. Moreover, ERβ, recently
discovered and cloned from rat prostate 264, was found to be expressed in many other tissues, including
injured arteries 265; 266. Iafrati et al. 267 reported that the prevention of medial enlargement (ie, smooth muscle
cell proliferation) by estradiol is preserved in ERα-deficient mice in a model of endovascular carotid injury.
The same group also reported that estradiol inhibits medial enlargement in the injured carotid of ERβ-
deficient mice 268. Recently published study demonstrated that ERα but not ERβ mediates the beneficial
effect of estradiol on re-endothelialization and potentially the prevention of atherosclerosis 269. These studies
suggest that re-endothelialization could be mediated specifically by the ERα gene, whereas arterial smooth
muscle cell proliferation could be mediated by a third, as yet unidentified ER gene.
Both, in vivo and in vitro studies have supported the view that estrogen influences vascular tone via an NO-
mediated mechanism. Basal NO release is greater in female rabbits than in male or oophorectomized female
rabbits 270. Estradiol restores impaired endothelium-dependent relaxation in the abdominal aorta of balloon-
injured and high cholesterol diet-treated rabbits, and also increased basal NO release 271. Inhibition of NOS
reduces the antiatherosclerotic effect of estrogen in cholesterol diet-induced atherosclerosis in the rabbit
aorta 272. This direct antiatherosclerotic effect of estrogen was present, absent, or reversed, depending on the
state of the arterial endothelium
273. In the undamaged aorta, cholesterol accumulation of the placebo rabbits
was significantly increased from week 4 to 8, and this increase was almost completely inhibited by estrogen.
In the balloon-injured aorta, the estrogen and placebo rabbits accumulated similar amounts of cholesterol in
the re-endothelialized areas, but in the de-endothelialized areas the estrogen group accumulated significantly
more cholesterol than the placebo group 273.
The above data suggest that estrogen may confer cardiovascular protection by improving the function of
vascular endothelium. In normotensive and hypertensive women, endogenous estrogen can prevent or
decrease the age-associated endothelial dysfunction 274; 275. In postmenopausal women, acute administration
of estrogens increases endothelium-dependent vasodilation of coronary arteries 276-278 and forearm
resistance vessels 279; 280. Acute estradiol also potentiated vasodilation induced by the endothelium-
independent vasodilator SNP 279; 280. Sublingual estradiol acutely increases peripheral blood flow 281 and may
282 or may not 283 have beneficial effects on exercise-induced myocardial ischaemia.
In cross-sectional studies effects of long-term HRT on endothelial function have been contradictory 284; 285, but
prospective clinical trials support the view that oral HRT increases flow-mediated vasodilation of the brachial
artery 286-289. Also, withdrawal of oral estrogen therapy is associated with a deterioration in endothelial
function in postmenopausal women with CHD 290. Recent findings also suggest that oral CEE of 0.625 mg
daily for 12 weeks improves forearm resistance artery endothelial function in postmenopausal women and
that this beneficial effect is even greater in subjects that are hypertensive 289. The effects of transdermal HRT
on endothelial function have been contradictory 252; 279; 291. Many of the previous studies have been
uncontrolled 279; 287; 290 or cross-over 286; 288 design trials. Estradiol concentrations were measured in some 252;
279; 288; 290; 291 but not in all 286; 287 previous studies. The oral estrogen preparation used in most 288; 290 but not in
all 286 studies has been CEE. Sequential oral or transdermal HRT for one year did not change circulating
markers of endothelial function or urinary excreted prostacyclin metabolites 292. In none of the previous
studies have effects of oral and transdermal estradiol on in vivo endothelium-dependent vasodilatation been
compared, and no study addressed the question of whether the acute effects of estradiol are responsible for
enhanced endothelial function during long-term therapy.
Concomitant progestin administration could counteract beneficial effects of estrogen on endothelium-
dependent vasodilatation. Three available human studies have reported conflicting data 252; 284; 285. It remains
therefore to be established whether progestin opposes effects of estrogen on endothelial function. This is an
important question considering the need to combine ERT with progestin for primary and secondary
prevention of CV disease.
Arterial stiffness
Increased pulse pressure, a measure of arterial stiffness, is an independent predictor of the incidence of
CHD and overall mortality among elderly 131. Little, however, is known regarding effects of HRT on arterial
stiffness. Menopause is suggested to be associated with the reduced elasticity of the carotid arteries in 50-
year-old women 293. Hayward et al. 294 recently examined whether acute sublingual estradiol administration
has direct hemodynamic effects on ventricular or vascular (aortic impedance and pulse wave velocity)
function. Despite achieving supraphysiological serum concentrations of estradiol (mean 3190±2216 pmol/l),
no significant acute effects of estradiol administration on ventricular or arterial function were found 294. There
24
was a significant decrease in cardiac output and heart rate in the pooled data of all subjects following
estradiol administration, which is surprising because a number of studies have shown that estrogen may act
directly as a smooth muscle relaxant 295; 296. Previous studies, which were noninvasive and did not measure
serum estradiol concentrations, have suggested that short-term estrogen has a positive inotropic effect and
increases peak aortic flow velocity 297; 298. Contrary to the study by Hayward et al. 294, Stefanadis et al. 299
showed increased aortic distensibility and decreased augmentation index following intravenous estradiol
administration.
Cross-sectional studies of effects of HRT on arterial stiffness have produced variable results. In some studies
300-302 postmenopausal HRT was associated with decreased carotid arterial stiffness, and in other studies
none of the measures of arterial stiffness differed between users and nonusers of HRT 303; 304. Regarding
intervention studies, Waddell et al. 305 measured systemic arterial compliance and pulse wave velocity in
postmenopausal women on and 4 weeks off HRT. Systemic arterial compliance decreased after
discontinuation of HRT, but this was not due to change in large artery compliance, because aorto-femoral
pulse wave velocity remained unchanged while pulse wave velocity in the femoral-dorsalis pedis region
increased significantly. Short-term (up to 6 months) oral 306 and transdermal 307 HRT have been shown to
reduce the pulsatility index , which is thought to reflect impedance to blood flow distal to the point of sampling,
in carotid artery. HRT has also been associated with a reduction in pulsatility index in peripheral arteries 308.
This could represent a direct effect of estrogen on the peripheral resistance arteries.
Long-term vascular effects of oral and transdermal HRT on carotid and uterine artery resistance to blood flow
have also been studied 309. Oral and transdermal sequential HRT were similarly effective at one year in
reducing impedance to flow 309. In none of the previous studies have effects of oral and transdermal estradiol
on preresistance and resistance vessel function in healthy postmenopausal women been compared.
Antioxidants
Oxidative modification of LDL has been implicated in the initiation and progression of atherosclerosis 310. It
has been proposed that LDL accumulates in the subendothelial space of lesion-prone arterial sites where
vascular cells oxidatively modify this LDL into a form that is internalized by scavenger receptors of resident
macrophages, resulting in the formation of ”foam cell” that is characteristic of early atherosclerotic lesions 311.
Emerging evidence suggests that estrogens may act as antioxidants. In vitro the antioxidant potency of
estradiol is 10-100 times greater than alpha- and gamma-tocopherol 312. Estrogens with a phenolic structure
protect LDL from both cellular and copper-mediated oxidation in vitro 313-316. One problem with these studies
is the need to use supraphysiological estrogen concentrations (>1µmol/l) to protect LDL from oxidation.
Shwaery et al. 317 showed an inhibition of in vitro oxidation of LDL in the whole serum, by lower
concentrations (0.1-100 nmol/l) of estradiol, while two other studies 318; 319 found physiological concentrations
of estradiol not to protect LDL from copper-mediated oxidation. There is some evidence that vitamins C and
E can enhance the antioxidant effect of estradiol by preventing LDL oxidation by copper ions or cells; thus
antiatherosclerotic activity of estradiol may occur at concentrations within the physiologic range 320; 321.
Effects of physiological doses of estradiol on oxidative modification of lipids and lipoproteins have been
inconsistent. Sack et al. 322 demonstrated that both acute intravenous estradiol and three weeks of
transdermal estradiol inhibited LDL oxidability. Also three weeks of oral CEE and combined HRT inhibited
lipid peroxidation 323. Guetta et al. 324 observed that combined administration of estradiol and vitamin E both
protect LDL oxidability. In a randomized crossover trial where either oral CEE or transdermal estrogen were
administered for one month, LDL oxidability remained unchanged 325.
Insulin resistance
Hormone replacement therapy and insulin sensitivity
Studies of the effect of the menopause on carbohydrate metabolism and insulin sensitivity have produced
variable results 228; 326; 327. In a study, where prospectively determined changes in coronary risk factors that
were attributable to natural menopause were followed, natural menopause did not adversely affect plasma
glucose or insulin levels 228. Cross-sectional comparison of estimates of insulin sensitivity in small number of
pre- and postmenopausal women reported a reduction in insulin sensitivity in postmenopausal women 326. A
recent analysis of the European Group for the Study of Insulin Resistance of the relationship between age
and insulin sensitivity found no difference in insulin sensitivity, measured using euglycemic insulin clamp
technique, between women with a mean age of 45 yr (n=76) and those aged 65 yr (n=88) 327.
25
Epidemiological evidence suggests a decreased incidence of type 2 diabetes in postmenopausal women on
estrogen therapy 328. Also, postmenopausal estrogen use has been associated with lower fasting plasma
glucose and insulin concentrations 173. In contrast to these epidemiological data, in clinical trials effects of
HRT on insulin sensitivity have more often been neutral or negative than positive. Effects of HRT on insulin
sensitivity have previously been examined in three placebo-controlled studies using the euglycemic insulin
clamp technique 329-331. No significant change in insulin sensitivity was observed after 6 weeks of transdermal
estradiol, or after addition of progestin for a further 6 weeks 329. Insulin sensitivity with the use of continuous
combined estradiol/norethisterone acetate preparations was measured by means of a two-step euglycemic
insulin clamp at baseline and after 3 months of therapy 331. A low dose estradiol (1 mg/day) did not change
insulin sensitivity, while a high dose of estradiol (2mg/day) slightly decreased insulin sensitivity 331.
Regarding the question of whether the route of estradiol administration is a possible factor in insulin
sensitivity, a comparison of effects of 1.25 mg oral CEE and 100 µg of transdermal estradiol found no effect
of either treatment on insulin sensitivity 330. Another study also reported no difference in oral glucose
tolerance 332. Godsland et al. 333 reported a slight, but significant deterioration in glucose tolerance with oral
(CEE 0.625 mg/day + levonorgestrel) compared to transdermal (50 µg/day + norethindrone acetate) HRT, but
no change in insulin sensitivity, as deduced from an iv glucose tolerance test. Studies using indirect
measures of insulin sensitivities have reported improvements both with transdermal 334-336 and oral 335; 337
estrogen preparations. However, in none of the previous studies have effects of unopposed oral and
transdermal estradiol on insulin sensitivity using euglycemic insulin clamp technique been compared.
Markers of inflammation, coagulation and fibrinolysis
Inflammation
Monocyte adhesion and subendothelial migration, which are in part mediated by adhesion molecules,
characterize initiation of the atherosclerotic process. The pathophysiologic relevance of soluble cell adhesion
molecules measured in human sera is supported by their localization in atherosclerotic plaques 338; 339. For
example, VCAM-1 expression with an associated leukocyte infiltration has been demonstrated in
microvessels within advanced human plaques 340. Furthermore, there is now ample evidence that advanced
plaques most frequently rupture at sites enriched with inflammatory cells and lipid 341. Thus, the effect of
estrogen in inhibiting VCAM-1 expression and monocyte adhesion and subendothelial migration in rabbits in
vivo could have impact on all stages of atherogenic process and could be considered antiatherogenic in
women before menopause 273; 342. In women with CHD, serum concentrations of E-selectin, intercellular
adhesion molecule-1 (ICAM-1), and VCAM-1 are higher in postmenopausal non-users than in users of HRT
343. Oral CEE has been reported to significantly reduce serum concentrations of E-selectin 288; 344; 345, ICAM-1
344, and VCAM-1 344 relative to pretreatment values. Higher concentrations of adhesion molecules have been
associated with risk of first myocardial infarction in two studies, where observed effects were independent of
lipid concentrations 346; 347.
Regarding CRP, individuals with existing CHD 348, and those at greatest risk of future CV events 349; 350 have
higher serum concentrations of CRP, reflecting up-regulation of at least some aspects of inflammation
involved with atherosclerosis. In a cross-sectional study of long-term elderly users (mean age 73 years),
hormone use was associated with over 50% higher CRP concentrations compared with nonusers 351. Later
this finding was observed in a group of women representative of those who have more recently started HRT
(mean age 52 years) 352. In the same study, both estrogen alone and combined with progesterone were
associated with increased CRP concentrations 352. In PEPI-trial, assignment to HRT compared with placebo
was associated with increased concentrations of CRP, which were maintained during 3 year follow-up 345.
Similarly, in a 12-week study of postmenopausal women initiating micronised estradiol alone or sequentially
combined with a progestin, CRP concentrations were found to increase 353. The clinical significance of these
data is uncertain but may have implications for women initiating HRT, especially, because elevated
concentrations of CRP are associated with increased CV risk among otherwise healthy women 354. It is
known that CRP derives mainly from the liver 355, but inflammated vascular wall can also produce it, but the
relative contribution of these sources to serum CRP level is not known. Whether transdermal and oral
estradiol have similar effects on markers of inflammation has not been studied. Also the biologic relevance
of these differences in effects on markers of inflammation remains to be determined in future vascular
studies and clinical trials.
Coagulation
26
Hemostasis is the result of interplay of four components: platelets, blood coagulation system, fibrinolytic
system, and endothelium. Figure 5 shows a diagram of the parts of the coagulation and fibrinolysis cascade.
The involvement of hemostatic system in atherogenesis is widely accepted 356-358, and enhanced blood
coagulation and decreased fibrinolysis may contribute to the formation of atherosclerotic lesions. Factor VII
plays a vital role in tissue factor-mediated coagulation 359. Recently developed assays specific for activated
factor VII 360 will help clarify its role. In previous cross-sectional studies, increases 219; 351; 361 and decreases
362 in factor VII concentrations amongst HRT users have been reported. In prospective trials the results have
also been conflicting since HRT has increased 254; 363 or not changed 364 factor VII concentrations. Studies
addressing clotting factors in conjuction with markers of intravascular coagulation have had conflicting results
with no net effect 364; 365 or an effect toward hypercoagulability 363; 366; 367. In only one of the previous studies
has unopposed estradiol been used 363, and in none of the previous studies have effects of oral and
transdermal estradiol on coagulation markers been compared.
Fibrinolysis
Fibrinolysis is the process which leads to fibrin degradation and removal of the hemostatic plug. Fibrinolysis is
activated by plasminogen activators, which cleave the inactive plasminogen to proteolytic enzyme plasmin.
Plasmin degrades fibrin stepwise to fibrin degradation products (Fig. 5.). Data concerning effect of HRT on
fibrinolysis seem rather consistent. In cross-sectional studies 351; 368; 369, current HRT users have had lower
plasminogen activator inhibitor-1 (PAI-1) antigen 351; 368; 369 and tissue-type plasminogen activator (tPA)
antigen 368; 369 concentrations than non-users. Two crossover trials found an almost 50% reduction in PAI-1
and tPA antigen levels from baseline after 6 weeks of oral CEE 363 or one month of oral CEE alone or
combined with progestin 325, while transdermal estradiol had no effect 325; 363. The decrease in PAI-1 antigen
concentrations was significantly correlated with an increase in D-dimer concentrations, suggesting that CEE
enhanced fibrinolysis 325. In a recently published prospective study, where oral CEE, simvastatin, and their
combination was used, only therapies including CEE decreased PAI-1 concentrations. Oral estradiol
combined with progestin 254; 367 has also been shown to decrease PAI-1 254; 367 as well as tPA 254 antigen
concentrations. Effects of oral and transdermal estradiol on markers of fibrinolysis have not so far been
compared.
27
TF
PL VIIa
Extrinsic tenase
TFPI TFPI
IX X
C1-inh.
XIa
XIIa HKa
XI
Contact
activation
PL
VIIIa
IXa
Intrinsic tenase
ATIII
Xa PL ATIII
Va
Prothrombinase
Prothrombin Plasminogen
F1+2
PL
TM Thrombin Plasmin
ATIII Antiplasmin
C1-inh.
APC
PC
Protein Case
Fibrinogen Fibrin Fibrin
degradation
products
FPA, FPB B-ß fragments
PAI-1 tPA
uPA
COAGULATION FIBRINOLYSIS
Figure 5. Blood coagulation and fibrinolysis (modified from ref.370). Extrinsic pathway of coagulation is
initiated when tissue factor (TF) is exposed from damaged endothelial cells or activated monocytes. TF and
activated factor VII (VIIa) on phospholipid (PL) surface form the extrinsic tenase complex which activates
both factor X and IX. Intrinsic tenase is formed by PL-bound VIIIa and IXa which has been activated by the
contact activation system or extrinsic tenase. Prothrombinase (Xa, Va on PL surface) catalyses thrombin
activation. Thrombin cleaves fibrinopeptides (FPA, FPB) from fibrinogen. Fibrin monomers are polymerized to
form insoluble fibrin. Thrombin in a complex with thrombomodulin (TM) activates also protein C (protein
Case). Activated protein C (APC) inhibits factors Va and VIIIa. Antithrombin III (AT III) is the major inhibitor of
coagulation. Fibrinolysis is started with activation of plasmin by tissue-type plasminogen activator (tPA). The
major inhibitor of fibrinolysis is plasminogen activator inhibitor 1 (PAI-1). Plasmin catalyses fibrin degradation.
HKa=activated high-molecular weight kininogen, TFPI=tissue factor pathway inhibitor, F1+2=prothrombin
fragment 1+2, uPA=urokinase-type plasminogen activator.
28
3. Aims of the study
The present studies were conducted to answer the following questions:
1) Is IFG associated with endothelial dysfunction? (Study I)
2) Does insulin therapy change in vivo endothelial function in patients with type 2 diabetes? (Study II)
3) Does ERT change endothelial function, and if so, is the change dependent on the route (oral vs.
transdermal) of estradiol administration? (Study III)
4) Does oral vs. transdermal ERT change insulin action on glucose metabolism, peripheral blood flow, or
arterial stiffness? (Study IV)
5) Does oral vs. transdermal ERT change markers of coagulation, fibrinolysis, inflammation and serum lipids
and lipoproteins? (Study V)
29
Table 4. Characteristics of the subjects.
STUDY I STUDY II STUDIES III-V
Type 2 diabetic Oral Transdermal
IFG Normal patients Normal estradiol estradiol Placebo
Number of subjects 17 12 18 27 9 11 7
Men/Women 17/0 12/0 14/4 20/7 0/9 0/11 0/7
Age 63±1 60±158±256±1 55±1 56±1 55±1
(years)
Body mass index 26.5±0.8 26.2±0.6 28.5±0.6 27.6±0.4 25.3±0.8 24.9±1.0 25.9±0.9
(kg/m2)
Systolic blood pressure 144±4 151±4 144±6 131±6 133±6 131±6 131±7
(mmHg)
Diastolic blood pressure 80±286±279±279±378±278±279±4
(mmHg)
Fasting plasma glucose 5.6±0.1 6.5±0.1 12.8±0.6*5.3±0.3 5.4±0.2 5.5±0.1 5.3±0.1
(mmol/l)
Fasting serum insulin 9±27±111±1*6±14±16±14±1
(mU/l)
Glycosylated hemoglobin 5.8±0.1 5.7±0.1 9.0±0.3*5.7±0.1 5.7±0.1 5.7±0.1 5.7±0.1
(%)
Serum total cholesterol 5.4±0.3 5.9±0.2 5.7±0.3 5.7±0.2 5.6±0.2 6.2±0.3 5.8±0.3
(mmol/l)
Serum LDL cholesterol 3.5±0.2 3.8±0.2 3.6±0.2 3.8±0.2 3.6±0.2 4.2±0.2 3.9±0.3
(mmol/l)
Serum HDL cholesterol 1.4±0.1 1.7±0.1 1.1±0.1*1.4±0.1 1.5±0.1 1.6±0.1 1.4±0.1
(mmol/l)
Serum triglycerides 1.2±0.1 1.1±0.1 2.1±0.3*1.4±0.2 1.4±0.2 1.0±0.1 1.3±0.2
(mmol/l)
Data are shown as mean±SEM. *P<0.05 for type 2 diabetic patients before insulin therapy vs. normal subjects.
30
4. Subjects and study designs
Baseline characteristics of the subjects are presented in Table 4. Aims and designs of the studies are listed
below. All studies were approved by local ethics committee.
I IFG and in vivo endothelial function
Aim: To determine whether endothelial dysfunction characterizes subjects with IFG, and if so, is it mediated
via mechanisms involving increased oxidative stress.
Design: Seventeen subjects with IFG and 12 subjects with normal fasting glucose concentrations participated
in the study. In vivo endothelial function (forearm blood flow responses to intrabrachially infused endothelium-
dependent and endothelium-independent vasodilators) was determined. The design of the endothelial
function test is shown in Fig. 6. Concentrations of antioxidants in serum or plasma (total radical trapping
trapping capacity (TRAP), uric acid, sulfhydryl groups, ascorbate, and serum FFA) were measured in blood
samples taken immediately before the endothelial function test.
1515
SNP
(µg/min)
310 7.515
ACh
(µg/min)
018304860
TIME (minutes)
Figure 6. Design of the endothelial function test used in protocols I-III. SNP and ACh were infused into the
brachial artery and the forearm blood flow responses were recorded. The numbers inside the white boxes
represent the doses of drugs, and the hatched areas denote times of the saline infusion before and between
the drugs.
II Effect of insulin therapy on endothelial function in type 2 diabetes
Aim: To determine whether insulin therapy changes in vivo endothelial function in patients with type 2
diabetes.
Design: Endothelial function (Fig. 6.) was measured in 18 type 2 diabetic patients previously treated with
metformin before and 6 months after combination therapy with bedtime human isophane insulin and
metformin. A group of 27 normal subjects was studied as a non-diabetic control group. The patients were
taught self-adjustment of the insulin dose on the basis of fasting plasma glucose measurements. The goal
was to lower fasting plasma glucose to ≤ 6.0 mmol/l and HbA1c to <7.5%. The patients visited the hospital
outpatient clinic monthly for 3 months after the start of insulin therapy and then 3-month intervals. We also
determined whether continued use of metformin affects endothelial function by studying 6 type 2 diabetics
chronically using metformin twice with a 6-month interval.
III Effect of oral and transdermal ERT on endothelial function in postmenopausal women (Figure 7)
Aim: To compare effects of oral and transdermal ERT on endothelial function and to determine whether these
changes are related to changes in serum free estradiol concentrations.
Design: Healthy postmenopausal women were recruited by a newspaper advertisement. The women were
screened by both the internist and gynecologist. Of the 36 subjects screened, a total of 27 women fulfilled the
inclusion criteria and were randomized to receive either oral (a daily 2 mg tablet, n=9) or transdermal (a patch
delivering 50 µg/day, n=11) estradiol or placebo (n=7). Endothelial function tests (Fig. 6.) were performed
after 0, 2, and 12 weeks of treatment. Serum total and free estradiol, estrone, sex hormone binding globulin
(SHBG) and serum lipid and lipoprotein concentrations were measured in blood samples taken immediately
before the endothelial function test.
31
1
5
1
5
0
T
IME (weeks)
T
RANSDERMAL ESTRADIOL 50 µg/day
O
RAL ESTRADIOL 2 mg/day
Endothelial
function test
P
LACEBO
212
Euglycemic
hyperinsulinemic clamp
Lipid and lipoproteins
Markers of inflammation,
coagulation and fibrinolysis
Hormone concentrations XX X
Figure 7. Design of the studies III-V.
IV Effect of ERT on insulin sensitivity of glucose metabolism and preresistance and resistance vessel
function in healthy postmenopausal women (Figure 8)
Aims: We hypothesized that ERT might enhance actions of insulin on pre-resistance or resistance vessels
and therefore compared effects of oral and transdermal ERT on basal arterial stiffness, peripheral blood flow,
and vascular resistance The actions of insulin on peripheral blood flow and resistance, arterial stiffness, and
glucose metabolism before and after 12-weeks of ERT were measured.
Design: All 27 women from Study III participated in this study. In each subject, an euglycemic
hyperinsulinemic clamp was performed at baseline and after 12 weeks of ERT. Endothelial function tests and
euglycemic hyperinsulinemic clamps were performed in separate days in each subject. Before and during
hyperinsulinemia, forearm blood flow (venous occlusion plethysmography), the augmentation index (pulse
wave analysis) and peripheral vascular resistance were determined as shown in Fig. 8. Because results
regarding insulin actions on blood flow, the augmentation index or glucose metabolism were not dependent
on the route of estradiol administration, results from the transdermal and oral estradiol groups were pooled.
V Effects of oral and transdermal ERT on markers of coagulation, fibrinolysis, inflammation and
serum lipids and lipoproteins in healthy postmenopausal women (Figure 7)
Aim: To compare effects of oral and transdermal estradiol on markers of coagulation, fibrinolysis, and
inflammation. We also wished to characterize effects of the route of estradiol administration in detail on
lipoprotein metabolism.
Design: All 27 women from Study III participated in this study. Markers of coagulation, fibrinolysis and
inflammation (serum CRP and soluble E-selectin), serum lipid, lipoprotein, and apolipoprotein concentrations
were measured and LDL particle size was quantitated at 0, 2 and 12 weeks of ERT.
32
1
5
1
5
-1 2
TIME (hours)
0
1
1
mU/kg·min
2
0% GLUCOSE
Insulin
PWA X
X
X
X
X
X
X
PVR
PWA = Pulse wave analysis calculation of augmentation index
PVR = Peripheral vascular resistance = MAP
forearm blood flow
X
Figure 8. Design of the euglycemic hyperinsulinemic clamp study used in protocol IV. Before and during
hyperinsulinemia, forearm blood flow (venous occlusion plethysmography), the augmentation index (pulse
wave analysis) and peripheral vascular resistance were determined.
33
5. Methods
5.1. Vascular function
Endothelial function (Studies I-III)
Endothelial function was assessed in forearm resistance vessels after an overnight fast by measuring
forearm blood flow responses to intra-arterial infusions of endothelium-dependent (ACh) and -independent
(SNP) vasodilators. Current smokers refrained from smoking for 12 hours prior to the study. An indwelling
cannula was inserted in an antecubital vein for blood sampling. A 27 G unmounted steel cannula (Coopers
Needle Works, Birmingham, UK), connected to an epidural catheter was inserted into the left brachial artery.
All drugs were infused at a constant rate of 1 ml/min in the following sequence: normal saline, SNP
(Nitropress, Abbott Labs, North Chigaco IL) 3 (low dose) and 10 (high dose) µg/min, ACh (Miochol, OMJ
Pharmaceuticals, San Germán, P.R.) 7.5 (low dose) and 15 (high dose) µg/min. Each dose was infused for 6
min, and infusion of each drug was separated by infusion of normal saline for 18 min, during which blood flow
returned to basal values. Forearm blood flow was recorded for 10 s at 15 s intervals during the last 3 minutes
of each drug and saline infusion period. Forearm blood flow was recorded in both the infused (experimental)
and control forearms simultaneously with the use of strain-gauge plethysmography as described below. Total
flows (basal and ACh- or SNP-stimulated) (studies I-III) and increases in flows above basal (total minus basal
flow) were calculated (studies I and III). We did not calculate the ratio of blood flow between the experimental
and control forearm, because blood flows in the control arm were similar to those in the experimental arm
basally and throughout the studies (I-III). The percentage changes in blood flow responses induced by
treatment (studies II and III) were calculated from the mean±standard error of mean (SEM) of individual
changes, which were defined as: 100*(flow after - flow before)/flow before treatment. The doses of SNP and
ACh were chosen, because both the lower (3 and 7.5 µg/min) and higher (10 and 15 µg/min) doses produce
similar increases in blood flow in normal subjects, and are associated with impaired vasodilation in
hypercholesterolemic subjects 41.
Arterial stiffness (Study IV)
The technique of pulse wave analysis was used to determine central aortic pressure and the augmentation
index, a measure of large artery stiffness 371(Study IV). All measurements were made from the radial artery, with
the wrist slightly extended and supported on a pillow, by applanation tonometry with the use of a Millar tonometer
(SPC-301; Millar Instruments, Houston, TX). Data were collected directly into a desktop computer and
processed with SphygmoCor Blood Pressure Analysis System (BPAS-1; PWV Medical, Sydney, Australia),
which allows continuous on-line recording of the radial artery pressure waveform. The radial waveform was
assessed visually to ensure that artifacts from movement and respiration were minimized. Recording for pulse
wave analysis were made basally and every 30 minutes during insulin infusion. The mean of three
measurements, each consisting of 15-20 sequentially recorded radial artery waveforms, was used to calculate
augmentation and the augmentation index as well as other parameters at a given time point. The integral system
software was used to calculate an average radial artery waveform, and to generate the corresponding ascending
aortic pressure waveform using a previously validated transfer factor 371-373. The aortic waveform was then
subject to further analysis for calculation of aortic augmentation, the augmentation index and central blood
pressure. The augmentation index was defined as the ratio of the difference between central aortic second and
first pressure peaks to pulse pressure. As suggested by O’Rourke and Gallagher 371, the radial blood pressure
was calibrated against the sphygmomanometrically determined brachial pressure, ignoring the small degree of
amplification between the brachial and radial sites. Reproducibility of this method accords with that reported by
other workers using different methodologies 108-110. With the wide range of augmentation index from -15.0 to
+45.0%, the within-observer difference was 0.49±5.37% and between-observer difference 0.23±3.80% 108. In
our own study, the coefficient of variation of augmentation index during sequential saline infusions lasting 6
hours averaged 5±1% 110.
5.2. Insulin action
Glucose metabolism (Study IV)
Whole body glucose uptake was quantitated using the euglycemic hyperinsulinemic clamp technique 374(Study
IV). The study was performed after an overnight fast starting at 7.30 a.m. Serum free insulin concentration was
increased to a predetermined level using a primed-continuous insulin infusion (Actrapid Human, Novo Nordisk,
Copenhagen, Denmark) (rate of continuous infusion 1 mU/kg⋅min for 120 min). Insulin and glucose were infused
34
into a 18 gauge catheter inserted in the left antecubital vein. Another 18 gauge catheter was inserted
retrogradely in the heated dorsal hand vein. This hand was kept in a heated chamber (65°C) to arterialize
venous blood 375. Normoglycemia was maintained by adjusting the rate of 20 % glucose infusion based on
plasma glucose concentrations performed at 5 minute intervals. Whole body glucose uptake was calculated
from the glucose infusion rate after correction for changes in the glucose pool size 374. Blood samples were
taken at 30 min intervals for measurement of the serum free insulin concentrations. We chose not to measure
hepatic glucose production during the clamp studies, because hepatic glucose production is completely
suppressed in normal subjects during a 1 mU/kg⋅min insulin infusion 376.
Resistance arteries (Study IV)
Forearm blood flow was measured with venous occlusion plethysmography using a mercury in silastic rubber
strain-gauge apparatus (Model EC-4, Hokanson, Bellevue, WA). The gauge was attached around the widest,
most muscular segment of the forearm. Because hand blood flow is predominantly directed through the skin,
and there is a high proportion of arteriovenous shunts, before the blood flow measurements, circulation to the
hand was interrupted by inflating a pediatric cuff around the wrist to above the systolic blood pressure. Venous
return was then occluded by a rapid cuff inflator (Rapid Cuff Inflator model E20, Hokanson) by increasing
pressure in a sphygmomanometer cuff around the upper arm to 40 mmHg. An analog-to-digital converter
(MacLab/4e, AD Instruments, Castle Hill, Australia) connected to a personal computer was used for recording
forearm blood flow. At least five flow curves were recorded for each flow measurement. Arterial inflow was
determined by drawing a tangential line for the first few pulses following inflation of the sphygmomanometer cuff.
The slope of this line reflects the volume change per unit time. Calibration was performed using the built-in
electronic calibration signal for a 1 per cent volume change, the height of which was used for blood flow
calculations. Means of the final five measurements of each recording period were used for analysis. The
reproducibility of this method was previously assessed by performing forearm blood flow measurements once
an hour during the 7-hour saline infusion 133. The coefficient of variation of these measurements averaged
13%, which is in line with one previous study, where the mean coefficient of variation for resting forearm
blood flow was 10.5% 377.
Large arteries (Study IV)
The technique of pulse wave analysis was used to determine central aortic pressure and the augmentation
index, a measure of large artery stiffness 371 (Study IV). Recording for pulse wave analysis was made basally
and every 30 minutes during insulin infusion as described in detail above.
5.3. Cardiovascular risk markers
Markers of coagulation and fibrinolysis (Study V)
Fibrinogen was determined as described by Clauss 378, using the ILTM Test Fibrinogen-C kit with the ILTM Test
calibration plasma (Instrumentation Laboratory Company, Lexington, MA). Factor VIII activity (FVIIIc) was
measured by a chromogenic method using the Coamatic R FVIII kit (Haemochrom Diagnostica AB, Mölndal,
Sweden). Factor VII antigen (FVIIag) was assayed with the Factor VII EIA kit (Dako A/S, Glostrup, Denmark,
a kind gift from Dr. Mirella Ezban (Novo Nordisk A/S, Målöv, Denmark). Activated factor VII (FVIIa) was
determined by the one-stage clotting method that uses soluble recombinant truncated tissue factor specific
for expression of FVIIa cofactor activity, a kind gift from Prof. James H Morrisey, Oklahoma Medical
Research Foundation, OK. Purified recombinant FVIIa (a kind gift from Dr. Mirella Ezban) was used as a
standard. F1+2 was measured with a sandwich-type of enzyme immunoassay (EnzygnostTM Micro from Dade
Behring Marburg GmbH, Marburg, Germany). Von Willebrand factor antigen (vWFag) was assayed by an in-
house enzyme immunoassay with antibodies from Dako A/S and Reference Plasma 100% (Immuno AG,
Vienna, Austria) as standard. PAI-1 activity (ChromolizeTM PAI-1), PAI-1 antigen (PAI-1ag, TintElize PAI-1)
and tPA antigen (tPAag, TintElize tPA) were determined with kits from Biopool (Umeå, Sweden). Plasmin-
antiplasmin complex (PAP) (EnzygnostTM PAP Micro) and D-dimer (EnzygnostTM D-dimer Micro) were
measured with kits from Dade Behring Marburg GmbH.
Marker of inflammation (Study V)
Serum CRP concentration was measured immunochemically (N Latex CRP mono, Dade Behring Marburg
GmbH).
35
Marker of endothelial activation (Study V)
Serum soluble E-selectin concentration was measured using Sandwich ELISA with a kit from R&D Systems,
Minneapolis, USA.
Serum lipid and lipoprotein concentrations (Studies I-V)
Serum lipoproteins were isolated by ultracentrifugation 379 using the following densities (d):VLDL d<1.006
g/ml, IDL d=1.006-1.019 g/ml, LDL d=1.019-1.063 g/ml, HDL d=1.063-1.210 g/ml. The concentrations of
cholesterol and triglycerides in serum and lipoprotein subfractions were determined by enzymatic colorimetric
assays (Hoffman-La Roche, Basel, Switzerland) using an autoanalyzer (Cobas Mira, Hoffman-La Roche).
Commercial kits were also used to measure the concentrations of phospholipids (WakoChemicals, Neuss,
Germany) and free cholesterol (Boehringen Mannheim, Mannheim, Germany) in lipoprotein subfractions.
Protein concentrations in the lipoprotein fractions were measured by the method of Kashyap et al. 380. Serum
FFA were quantified by the fluorometric method of Miles et al. 381. Serum Lp(a) concentrations were
determined by the Pharmacia Apolipoprotein (a) RIA assay system. Serum apo B was determined using an
immunochemical assay (Orion Diagnostica, Espoo, Finland), and serum apo A-I and apo A-II were
determined by a turbidimetric assay (Boehringer Mannheim). Apo E phenotyping was done from serum by
isoelectric focusing 382.
Quantitation of LDL particle size (Study V)
Non-denaturing polyacrylamide gel electrophoresis was performed on serum samples, stored at -80°C, using
gels cast in our laboratory, as described in detail 383. Gels were stained with Sudan Black B lipid stain and
scanned with a computer-assisted laser scanning densitometer (Personal Densitometer, Molecular
Dynamics, Sunnyvale, CA) using a 50 nm pixel size and 12-bit signal resolution. The particle diameter of the
major LDL peak was determined by comparing the mobility of the sample with the mobility of two reference
LDL preparations run of each gel. The particle diameters of the reference LDL preparations were determined
by electron microscopy 383. The inter- and intra-gel coefficients of variation for the control sample were 2.0
and 1.2 %, respectively.
5.4. Other measurements
Hormone concentrations (Studies I-V)
Serum follicle stimulating hormone (FSH) and SHBG (AutoDELFIATM SHBG, Wallac, Turku, Finland)
concentrations were determined by fluoroimmunometric assays. Serum estradiol (Estradiol-2, Sorin
Biomedica, Saluggia, Italy) , estrone (Estrone-RIA, Bühlman, Schönenbuch, Switzerland) and testosterone
(Spectria, Orion Diagnostica, Espoo, Finland) concentrations were measured by radioimmunoassays. Since
estradiol is bound to SHBG, and oral but not transdermal estradiol increases SHBG concentrations 384, free
estradiol concentrations were calculated as described by Dunn et al. 385 and free testosterone concentrations
as described by Anderson et al. 386. Serum free insulin concentrations were determined by double antibody
radioimmunoassay (Pharmacia Insulin RIA kit; Pharmacia, Uppsala, Sweden) after precipitation with
polyethylene glycol 387.
Measurement of plasma TRAP and water-soluble antioxidants (Study I)
TRAP was determined spectrophotometrically using a recently validated method 388. This assay uses 2´,7´-
dichlorofluorescein diacetate (DCFH-DA) to follow formation of free radicals during decomposition of 2,2´-
diazobis-(2-amidinopropane) dihydrochloride (AAPH). The DCF formation was monitored at 504 nm using a
Multiskan spectrophotometer (Labsystems, Helsinki, Finland). Plasma was mixed with PBS to a final dilution
of 1 %, and DCFA-DA was added (final concentration 14 µmol/l). The reaction was started by adding AAPH
(final concentration 56 mmol/l). Plasma ascorbic acid concentrations were measured using the
spectrophotometric method of Denson and Bowers 389. Plasma protein-bound thiols (sulfhydryl [SH] groups )
were determined as described by Ellman 390. Plasma uric acid concentrations were measured by an
enzymatic colorimetric assay (Roche Unimate 5 UA, Roche, Basel, Switzerland).
Metabolic and physical parameters (Studies I-V)
36
Plasma glucose concentrations were measured in duplicate with the glucose oxidase method 391, using the
Beckman Glucose Analyzer II (Beckman Instruments, Fullerton, CA). Glycosylated hemoglobin (HbA1C) was
measured by HPLC using a fully automated Glycosylated Hemoglobin Analyzer System (BioRad, Richmond,
CA). Urinary albumin excretion was measured using Micral-Test II strips (Boehringer Mannheim, Mannheim,
Germany) (I) or by an immunoturbidimetric (Hitashi Ltd) method with the use of an antiserum against human
albumin (Orion Diagnostica) (II). Blood pressure was measured with a mercury sphygmomanometer. Pulse
pressure was calculated from the difference between systolic and diastolic blood pressures. Mean arterial
pressure was determined by adding one third of the pulse pressure to diastolic blood pressure. Forearm
vascular resistance was calculated by dividing mean brachial artery pressure by forearm blood flow. Fat free
mass and the % body fat were determined using bioelectrical impedance analysis (Bio-Electrical Impedance
Analyzer System, Model #BIA-101A, RJL Systems, Detroit, MI) 392.
5.5 Statistical analyses
Analysis of group, time, and group x time effects between study groups were made using analysis of variance
(ANOVA) for repeated measures followed by pairwise comparison using the paired Student’s t test and group
comparisons using the unpaired Student’s t test followed by the Bonferroni correction when appropriate.
Correlation analyses were performed using Spearman’s nonparametric correlation coefficient. All calculations
were made using the SYSTAT software (Systat Inc., Evanston, IL). Log-transformation was used for
parameters that were non-normally distributed. All data are expressed as mean ± SEM. P-values less than
0.05 were considered to be statistically significant.
37
6. Results
6.1. Endothelial function in subjects with impaired fasting glucose (Study I)
Characteristics of the subjects (Table 4 and Table 1 of original publication I)
The subjects meeting criteria for IFG and subjects with normal fasting plasma glucose concentrations were
comparable with respect to physical (age, body mass index, and mean arterial pressure) and biochemical
parameters (HbA1C, fasting serum insulin concentrations, serum total, LDL, HDL cholesterol concentrations,
serum triglyceride and FFA concentrations) other than fasting plasma glucose concentrations. First-degree
relatives of subjects with IFG had significantly more often type 2 diabetes than did relatives of subjects with
normal fasting glucose concentrations (7 vs. 1 for IFG vs. normal, P<0.05). Serum and plasma antioxidant
concentrations (plasma TRAP, sulfhydryl groups, ascorbate, and serum uric acid) were comparable between
the groups.
Endothelial function
Blood flow responses to SNP and ACh are shown in Figure 9. Comparison of blood-flow responses during
the different drug doses showed that the responses to the low (5.9±0.7 vs. 10.9±1.3 ml/dl⋅min, IFG vs. normal
group, P<0.05) and high (9.1±1.2 vs. 13.2±1.5 ml/dl⋅min, respectively, P<0.05) doses of ACh were
significantly and by 46% and 31% lower in the IFG than the normal group, respectively (Fig. 9.). This was
also the case when data were expressed as the % increase in ACh-stimulated blood flow above basal (Fig.
10.). There were no differences in the blood flow responses to SNP between the groups (Fig. 9.). The ratios
of endothelium-dependent to -independent blood flows during the low and high dose drug infusions were 40%
and 30% lower in the IFG than in the normal groups (P<0.05). In all subjects, fasting plasma glucose (r=-
0.48, P<0.01) and HbA1c (r=-0.42, P<0.05) were significantly inversely correlated with blood flow during the
low-dose but not high-dose ACh infusion. Fasting plasma glucose was also significantly inversely correlated
with the ratio of endothelium-dependent and -independent blood flow during both the low-dose (Fig. 11.) and
high-dose (r=-0.40, P<0.05) infusions. None of the other parameters such as weight, body mass index, mean
arterial pressure, fasting serum insulin, LDL cholesterol and triglyceride concentrations correlated with blood
flow during intrabrachial infusions of endothelium-dependent or -independent vasodilators.
0 3 10
0
5
10
15
SNP ACh
07.5 15
*
**
IFG
Normal
Forearm blood flow
(ml/dl⋅min)
Figure 9. Forearm blood-flow responses to intra-arterial SNP and ACh infusions in the IFG and the normal
groups in the experimental (
) and control (----) arms.
∗
P<0.05 and
∗∗
P<0.01 for IFG vs. normal group.
38
ACh 7.5 µg/min
Normal IFG
0
500
1000
**
ACh 15 µg/min
Normal IFG
0
500
1000 *
Figure 10. ACh-induced % increases in forearm blood flow during low-dose and high-dose infusions.
∗
P<0.05 and
∗∗
P<0.01 for IFG vs. normal group.
80 100 120 140
0
1
2r=-0.53
p<0.01
Ratio of ACh 7.5/ SNP 3
Fasting plasma glucose (mg/dl)
Figure 11. The relationship between the fasting plasma glucose concentration and the ratio of endothelium-
dependent to -independent flow during infusions of ACh (7.5
µ
g/min) and SNP (3.0
µ
g/min).
6.2. Effect of insulin therapy on endothelial function (Study II)
Characteristics of the subjects (Table 4)
The type 2 diabetic patients had higher fasting plasma glucose, HbA1c, free insulin, and total triglyceride
concentrations and lower HDL cholesterol concentrations compared to normal subjects. Insulin therapy
significantly decreased concentrations of fasting plasma glucose, HbA1c, total triglycerides, and FFA, and
increased concentrations of free insulin.
Endothelial function
Blood flow responses to SNP and ACh before and after insulin therapy and in normal subjects are shown in
Figure 12. Before insulin therapy, blood flow during infusion of the low (6.7±0.6 vs. 9.3±0.8 ml/dl⋅min, p<0.05)
and high (7.5±0.7 vs. 11.6±0.9 ml/dl⋅
⋅⋅⋅min, p<0.01) doses of ACh were significantly blunted in the type 2
diabetic patients compared to the normal subjects. Insulin therapy significantly increased the blood flow
response to the high dose of ACh by 44% (7.5±0.7 vs. 10.8±1.6 ml/dl⋅min, before vs. after, p<0.05). Forearm
blood flow responses to both the low (7.8±0.4 vs. 9.1±0.4 ml/dl⋅min, p<0.05) and high (11.0±0.8 vs 13.0±0.
39
0 3 10
0
5
10
15
Experimental arm
Control arm
SNP ACh
Normal subjects, n=27
After insulin therapy
07.5 15
Type 2 diabetic patients, n=18
Before insulin therapy
*
*
+++
*
FOREARM BLOOD FLOW (ml/dl⋅
⋅⋅⋅min)
Figure 12. Forearm blood flow responses to intra-arterial SNP and ACh infusions in type 2 diabetic patients
before (
g
) and after (
~
) insulin therapy and in normal subjects (
i
).
∗
P<0.05 for type 2 diabetic patients
before vs. after insulin therapy. +P<0.05 and ++P<0.01 for type 2 diabetic patients vs. normal subjects.
ml/dl⋅min, p<0.05) doses of SNP also increased significantly during insulin therapy. After insulin therapy,
blood flows during SNP and ACh infusions were not significantly different from those in normal subjects. No
correlation was found between changes in metabolic parameters and the blood flow during ACh and SNP
infusions.
6.3. Effect of oral and transdermal estrogen replacement therapy on endothelial function (Study III)
Characteristics of the subjects (Table 4 and Table 1 of original publication III)
At baseline, the groups were matched for biological and menopausal ages, body weight, blood pressure, and
serum concentrations of FSH.
Hormone concentrations (Figures 13 and 14)
Serum total estradiol concentrations were below the limit of detection in the placebo group (<20 pmol/L) at 0,
2 and 12 weeks. In the oral estradiol group, serum total estradiol concentrations increased from <20 pmol/L
at baseline to 378±49 at 2 weeks and 423±45 pmol/L at 12 weeks (p<0.01). In the transdermal estradiol
group, serum total estradiol concentrations increased from <20 pmol/L at baseline to 156±26 at 2 weeks and
216±31 pmol/L at 12 weeks (p<0.001). Serum total estradiol concentrations were significantly higher in both
the transdermal (p<0.05 and p<0.01 at 2 and 12 weeks vs placebo) and oral estradiol (p<0.001 at 2 and 12
weeks vs placebo) than in the placebo group, and higher in the oral than the transdermal estradiol group at 2
(p<0.001) and 12 (p<0.01) weeks (Fig. 13.). Serum SHBG concentrations remained unchanged in the
placebo and transdermal estradiol, but increased in the oral estradiol group by 133% from 72±11 to 168±11
nmol/L at 12 weeks (p<0.001, Fig. 14.). Due to the increase in serum SHBG concentrations, free estradiol
concentrations were similar in the oral (3.17±0.36 pmol/L) and transdermal (3.09±0.49 pmol/L) groups at 12
weeks. (Fig. 14.). Serum estrone concentrations remained unchanged in the placebo and transdermal
estradiol groups (Fig. 14.) but increased over 10-fold in the oral estradiol group from 238±34 at baseline to
2727±316 at 2 weeks and 2947±421 pmol/L at 12 weeks (p<0.001). Serum total testosterone concentrations
remained unchanged in the oral (0.99±0.02 vs 0.74±0.05 nmol/L, 0 vs 12 weeks), transdermal (1.13±0.17 vs
1.23±0.13, respectively) and placebo (1.07±0.12 vs 1.01±0.09, respectively) groups. Also due to increase in
40
Serum estradiol
0 4 8 12
0
250
500
***
**
*
***
**
**
pmol/l
Serum estrone
0 4 8 12
0
1000
2000
3000
4000
ORAL
TRANS
CONT
pmol/l
***
***
Endothelium-dependent
vasodilatation
0 4 8 12
0
100
200
**
Time (weeks)
% change flow response
to ACh 15 µg/min
Figure 13. Serum estradiol (top) and estrone (middle) concentrations at 0, 2 and 12 weeks. Bottom, %
change in forearm blood flow during infusion of ACh (15
µ
g/min) at 2 and 12 weeks vs. 0 week. Top and
middle: P<0.001 (ANOVA for repeated measures for oral vs. transdermal and placebo groups); bottom,
P<0.01 (ANOVA for repeated measures for oral vs. transdermal and placebo groups).
∗
P<0.05,
∗∗
P<0.01
and
∗∗∗
P<0.001 for pairwise comparisons between groups.
serum SHBG concentrations, serum free testosterone concentrations decreased in the oral estradiol group
from 11±1 at baseline to 4±1 pmol/L at 12 weeks (p<0.001), and remained unchanged in the transdermal
(14±2 vs 14±2 pmol/L, respectively) and placebo (15±2 vs 14±2 pmol/L, respectively) groups.
41
CONT TRANS ORAL
0
100
200
nmol/l
CONT TRANS ORAL
0
100
200 *** ***
nmol/l
CONT TRANS ORAL
0
1
2
3
4
pmol/l
CONT TRANS ORAL
0
1
2
3
4*** ***
pmol/l
CONT TRANS ORAL
0
1000
2000
3000
4000
pmol/l
CONT TRANS ORAL
0
1000
2000
3000
4000
******
pmol/l
Serum Free Estradiol
Serum SHBG
Serum Estrone
0 weeks
0 weeks
0 weeks
12 weeks
12 weeks
12 weeks
Figure 14. Serum SHBG (top), free estradiol (middle), and estrone (bottom) concentrations at 0 and 12
weeks.
∗∗∗
P<0.001.
Endothelial function
In the oral estradiol group, total (basal and ACh-stimulated) flow during infusion of the low dose of ACh
increased from 7.6±0.9 at 0 weeks to 8.9±0.9 at 2 weeks and by 92±26 % to 13.0±1.1 ml/dl⋅min at 12 weeks
(p<0.01 vs 0 and 2 weeks). The results were similar when total flows during infusion of the high dose of ACh
were used in the analysis as shown in Figure 15. The % increases in flow above basal compared to 0 weeks
averaged 21±14% at 2 weeks and 120±34% at 12 weeks during the low dose ACh infusion, and 22±10% and
119±46%, respectively, during the high dose ACh infusion. In the same group, total flow during infusion of the
low dose of SNP averaged 8.2±0.9 at 0 and 9.3±0.8 at 2 weeks (NS vs 0 weeks) and 11.6±0.7 ml/dl⋅min at 12
weeks (p<0.05 vs 0 weeks). Similarly, total flow during infusion of the high dose of SNP increased
significantly at 12 weeks compared to baseline (Fig. 15.) The % increases in flow above basal compared to 0
weeks averaged 16±12% and 50±18% at 2 and 12 weeks during the low dose SNP infusion and 28±12% and
42
64±19% during the high dose SNP infusion. These results remained unchanged if the flow above basal
during infusion of the low and high dose of ACh and SNP were used in data analysis. In the transdermal and
placebo groups, there were no significant changes in the blood flow responses to any of the drugs. None of
the hormone concentrations correlated with measures of blood flow within the individual groups.
0 2 12
9
12
15
18
*
02 12
9
12
15
18
*
02 12
0
50
100
150
**
Time (weeks)
02 12
0
50
100
150 Oral
Transdermal
Placebo
*
Time (weeks)
Figure 15. Forearm blood flows (top) and % changes in flows at 2 and 12 weeks vs. 0 week during infusions
of ACh (15
µ
g/min) and SNP (10
µ
g/min).
∗
P<0.05 and
∗∗
P<0.01.
6.4 Effect of estrogen replacement therapy on insulin sensitivity of glucose metabolism and
preresistance and resistance vessel function in healthy postmenopausal women (Study IV)
Peripheral blood flow and vascular resistance (Fig. 16.)
Basal forearm blood increased significantly and by 25±8% in the estradiol but not in the placebo group (P<0.01)
(Fig. 16.). In the estradiol group, brachial diastolic blood pressure also decreased slightly from 78±2 to 75±2
mmHg (P<0.05) as did peripheral vascular resistance (p<0.01) in the estradiol but not in the placebo group (Fig.
16.). Forearm blood flow increased similarly in the women using oral (1.4±0.1 vs. 1.8±0.2 ml/dl⋅min, 0 vs 12
weeks, p<0.05) and transdermal (1.6±0.1 vs. 1.9±0.1 ml/dl⋅min, 0 vs. 12 weeks, p<0.05) estradiol. Also,
peripheral vascular resistance decreased similarly in the women using oral (63±5 vs. 50±3 mmHg/(ml/dl⋅min), 0
vs. 12 weeks, p<0.05) and transdermal (68±4 vs 55±6 mmHg/(ml/dl⋅min), 0 vs. 12 weeks, p<0.05) estradiol.
Basal brachial systolic blood pressure or pulse pressure did not change in the estradiol or placebo groups.
Central hemodynamic parameters
Aortic diastolic blood pressure decreased significantly (79±2 vs. 76±2 mmHg, 0 vs. 12 weeks, p<0.05) in the
estradiol but not in the placebo (81±4 vs 79±3 mmHg, 0 vs 12 weeks, NS) group. Augmentation, the
augmentation index, and aortic systolic blood pressure remained unchanged in the estradiol and placebo
groups.
43
0303
0.0
2.5
5.0
7.5
Time (months)
M-Value (mg/kg BW⋅min)
0303
0.0
0.5
1.0
1.5
2.0
2.5
**
Time (months)
Forearm blood flow
(ml/dl⋅min)
0303
50
60
70
80
90
100 Estradiol
Placebo
*
Time (months)
Diastolic blood pressure
(mmHg)
0303
0
25
50
75 **
Time ( months)
Peripheral vascular
resistance
(mmHg/ml/dl⋅min)
Figure 16. Basal brachial diastolic blood pressure, basal forearm blood flow, whole body insulin sensitivity (M-
value), and basal peripheral vascular resistance in the estradiol and placebo groups at 0 and 12 weeks.
∗
P<0.05
and
∗∗
P<0.01. BW=body weight.
Insulin action on central (Fig. 17.) and peripheral hemodynamic parameters
Augmentation was significantly acutely decreased by insulin within 30 minutes in the estradiol group both at 0
and 12 weeks (Fig. 17.). The decrease in augmentation by insulin was not altered by estradiol treatment (Fig.
17.). The augmentation index also significantly decreased by insulin within 30 minutes both at 0 and 12 weeks
(Fig. 17.). The ability of insulin to decrease the augmentation index was not altered by treatment in any group
(Fig. 17.). Acute hyperinsulinemia did not change heart rate (62±2 vs. 64±1 beats/min, basal vs. 30-120 at 0
weeks and 60±1 vs 61±1 beats/min, basal vs. 30-120 min at 12 weeks or ejection duration (338±4 vs. 337±3
ms, basal vs. 30-120 min at 0 weeks and 349±4 vs 340±3 ms, basal vs. 30-120 min at 12 weeks) in the estradiol
group. Estradiol therapy also did not change heart rate (64±1 and 61±1 beats/min, 0 and 12 weeks, NS) or
ejection duration (337±3 and 340±3 ms, respectively, NS) during hyperinsulinemia.
Insulin action on glucose metabolism (Fig. 16.):
Fasting plasma glucose (5.5±0.1 and 5.4±0.1 mmol/l vs. 5.3±0.1 and 5.6±0.1 mmol/l), HbA1C (5.8±0.1 and
5.6±0.1% vs. 5.7±0.1 and 5.7±0.1%), and serum free insulin concentrations (29±3 and 25±3 pmol/L vs. 26±6
and 28±4 pmol/L remained unchanged in the estradiol at 0 and 12 weeks vs. placebo group at 0 and 12 weeks.
During the insulin infusion serum free insulin concentrations averaged at 0 weeks 439±12 (30-120 min) and at
12 weeks 407±17 pmol/L in the estradiol group (NS), and 447±22 and 435±26 pmol/L in the placebo group,
respectively (NS). During hyperinsulinemia plasma glucose concentrations were maintained at 4.8±0.3 mmol/l
(30-120 min) at 0 weeks and 5.1±0.1 mmol/l at 12 weeks in the estradiol group and at 5.1±0.1 and 5.1±0.1
mmol/L in the placebo group, respectively (NS). Whole body insulin sensitivity remained unchanged in the
estradiol and placebo groups (Fig. 16).
44
030 60 90 120
10.0
12.5
15.0
17.5 0 months
3 months
*
**
***
**
*
AUGMENTATION (mmHg)
025 50 75 100 125
25
30
35 0 months
3 months
TIME (min)
AUGMENTATION INDEX
(%)
*** **
** *** ***
Figure 17. Augmentation and augmentation index (augmentation/pulse pressure) before and during insulin
infusion in the estradiol group at 0 and 12 weeks.
∗
P<0.05,
∗∗
P<0.01, and
∗∗∗
P<0.001 for change in
augmentation and augmentation index at a given time point vs. 0 min.
45
6.5. Effects of oral and transdermal estrogen replacement therapy on markers of coagulation,
fibrinolysis, inflammation and serum lipids and lipoproteins in healthy postmenopausal women (Study
V)
Markers of coagulation (Fig. 18)
Oral estradiol therapy significantly increased factor VII activity, factor VII antigen, and prothrombin fragment
1+2 concentrations (p<0.05) (Fig. 18.). Other markers of coagulation remained unchanged with no significant
differences between treatment groups.
0 2 12
450
500
550
600
650
Oral estradiol
Transdermal
Placebo
*
Factor VII antigen
(ng/mL)
0 2 12
0.75
1.00
1.25
1.50
1.75
*
Time (weeks)
Prothrombin fragment 1+2
(nmol/L)
Figure 18. Plasma concentrations of factor VII antigen (upper panel) and prothrombin fragment 1+2 (lower
panel) at 0, 2, and 12 weeks.
∗
P<0.05 (ANOVA for repeated measures for changes between groups).
Markers of fibrinolysis (Fig. 19.)
Oral estradiol therapy significantly lowered plasma tPA antigen, PAI-1 antigen concentrations, and PAI-1
activity (p<0.05), and markedly increased the concentration of D-dimer (Fig. 19.). Other markers of
coagulation remained unchanged with no significant differences between treatment groups.
46
0 2 12
5.0
7.5
10.0
Oral
Transdermal
Placebo
*
tPA antigen (ng/mL)
0 2 12
0
10
20
**
PAI-1 activity (IU/mL)
0 2 12
0
25
50
75
100
*
Time (weeks)
D-Dimer (µg/L)
Figure 19. Plasma concentrations of tissue-type plasminogen activator (tPA) (upper panel), plasminogen
activator inhibitor-1 (PAI-1) (middle panel), and D-Dimer (lower panel) at 0, 2, and 12 weeks.
∗
P<0.05
(ANOVA for repeated measures for changes between groups).
Markers of inflammation (Fig. 20.)
Serum CRP and soluble E-selectin concentrations remained unchanged in the placebo and transdermal
groups. In contrast, the concentration of serum CRP increased (p<0.05), while that of soluble E-selectin
decreased significantly in the oral estradiol group (Fig. 20.).
Serum lipids, lipoproteins and apolipoproteins
Serum total, VLDL and IDL triglycerides remained unchanged in all three groups. In the oral estradiol group,
LDL cholesterol decreased significantly within 2 weeks from 3.57±0.17 at baseline to 3.19±0.08 mmol/L
(p<0.05), and HDL cholesterol increased by 20 % (p<0.05 at 2 and 12 weeks). In addition, apo AI (p<0.05
and p<0.01 at 2 and 12 weeks) and apo AII (p<0.05 at 12 weeks) concentrations increased while those of
apo B (p<0.05 at 2 weeks) and Lp(a) (p<0.05 at 2 and 12 weeks) decreased significantly in the oral estradiol
group. In the transdermal and placebo groups, all lipid and lipo- and apoprotein concentrations remained
unchanged for the 12 week period.
47
0 2 12
0.5
1.5
2.5 Oral
Transdermal
Placebo
*
CRP (mg/L)
0 2 12
20
25
30
35
40
45
50
*
Time (weeks)
sE-Selectin (ng/ml)
Figure 20. Plasma concentrations of C-reactive protein (CRP) (upper panel) and soluble E-selectin (sE-
Selectin) (lower panel) at 0, 2, and 12 weeks.
∗
P<0.05 (ANOVA for repeated measures for changes between
groups).
48
7. Discussion
7.1 Endothelial function and altered glucose homeostasis
Impaired fasting glucose
Although in patients with type 2 diabetes, a defect in the vasodilatory response to ACh has been a rather
consistent finding 174-189, this study provides the first evidence that even mild fasting hyperglycemia is
associated with a similar defect. Regarding possible causes of this defect, acute hyperglycemia induced
endothelial dysfunction is unlikely, because in the IFG group average fasting plasma glucose concentration
was only 0.9 mmol/l higher than in the normal group. In previous studies addressing effects of acute
hyperglycemia on endothelial function results have been inconsistent 393-395. In two studies, maintenance of
glucose at 15 and 16 mmol/l for 24 and 7 hours with a continuous intra-arterial infusion of 5% glucose did not
change vascular function 393; 394 whereas in another study 6 hours of local hyperglycemia (17 mmol/l)
achieved by an intra-arterial infusion of 50% dextrose and systemic infusion of octreotide decreased
methacholine-induced vasodilatation 395. Increased oxidative stress has been suggested to cause endothelial
dysfunction in type 2 diabetes 396. In human aortic endothelial cells, high glucose increases production of free
radicals such as superoxide anion (02-)397. The latter could inactivate NO in the absence of adequate
antioxidant defense mechanisms thus affecting both the biotransformation of exogenous nitrates and
inactivation of endogenously released NO 398. In human plasma, vitamin C and protein thiols serve as first-
line defense mechanisms against oxidative stress although they explain only 23-28% of TRAP 181; 399; 400. In
the present study, serum concentrations of TRAP, vitamin C, and protein thiols were similar between the two
groups arguing against hyperglycemia induced endothelial dysfunction via increased oxidative stress. Also, if
rapid inactivation of NO by 02- occurred, it would be expected to also impair vasodilation by exogenous NO,
which was not observed in the present study.
Traditional causes of endothelial dysfunction amongst non-diabetic subjects include age, cholesterol
concentrations and hypertension. These parameters did not explain endothelial dysfunction in the subjects
with IFG. The IFG group also did not have features of the insulin resistance syndrome such as
hyperinsulinemia, hypertriglyceridemia, low HDL cholesterol or increased FFA concentrations. Whether the
IFG group had endotheliopathy before the development of insulin resistance as has been previously
suggested 401, or was predisposed to CV disease for other reasons, is unknown. The strong family history of
diabetes in the IFG group is consistent with both of these possibilities.
Type 2 diabetes
In study II, we demonstrated that patients with type 2 diabetes had an impaired response to the endothelium-
dependent vasodilator ACh, but a normal response to the endothelium-independent vasodilator SNP. This
finding is consistent with 13 out of 18 previous studies which also found endothelium-dependent
vasodilatation to be impaired in type 2 diabetes 174-177; 179-189; 191; 192; 402. The reason for normal endothelial
function in the study of Avogaro et al. 192 could be due to the small number of subjects studied (10 patients
with type 2 diabetes and 6 normal subjects). Regarding endothelium-independent vasodilatation, results have
been less consistent, since in 6 out of 18 studies this response has been impaired 174; 177; 179; 188; 189; 191 and in
11 of 18 studies normal 175; 176; 180-187; 192. In the current study, the effect of L-NMMA on basal and ACh
stimulated forearm blood flow was not determined. We have previously demonstrated that blood flow is
similar in type 2 diabetic patients and normal subjects during co-administration of L-NMMA with ACh implying
that the defect in endothelium-dependent vasodilation is due to defects in NO production or action 181. Normal
responses to verapamil in two studies 176; 179 suggest that a generalized abnormality of vascular smooth
muscles does not explain vascular dysfunction in patients with type 2 diabetes.
Classic CV risk factors including age, total and LDL cholesterol concentrations and hypertension did not
explain endothelial dysfunction in patients with type 2 diabetes, and none of the patients with type 2 diabetes
smoked. An atherogenic lipoprotein phenotype including increased plasma triglyceride concentrations, small
LDL size and low HDL cholesterol concentrations, and increased FFA concentrations 379; 403; 404, increased
oxidative stress 396, and hyperglycemia induced formation of advanced glycosylation end products (AGEs) 405
are all possible biochemical mediators of endothelial dysfunction in type 2 diabetes. Hypertriglyceridemia and
low HDL-cholesterol have been shown to be associated with endothelial dysfunction in men with diet treated
type 2 diabetes 177. The hypertriglyceridemia in type 2 diabetes is accompanied by increased production of
small, dense LDL particles, which are highly susceptible to oxidative modification 406 and are associated with
impaired endothelial function independent of other lipid parameters 64. Acute marked increases in circulating
49
FFA concentrations cause endothelial dysfunction 403; 407-409, whether chronic small elevations impair
endothelial function in patients with type 2 diabetes is unknown 210.
Treatment of endothelial dysfunction in patients with type 2 diabetes
During combination therapy with bedtime insulin and metformin endothelium-dependent and to a lesser
extent endothelium-independent vasodilatation improved. These changes were not observed during chronic
metformin therapy suggesting that the improved vascular function was due to insulin therapy. During insulin
therapy mean body weight, blood pressure, and the LDL cholesterol concentration remained unchanged.
Weight gain is an inevitable consequency of insulin therapy, if glycemic control improves, and averages 2 kg
for 1% decrease in HbA1C during treatment regimens including insulin alone 410. The reason for the lack of
significant weight gain could be due to at least two factors. First, the improvement in glycemic control was
relatively small (-1.4%). Second, we have previously shown that metformin effectively prevents weight gain
during insulin therapy 411. Weight gain in patients using insulin alone and those using insulin and metformin
suggest that the weight gain-sparing effect of metformin is due to reduced energy intake 412. Although the
UKPDS suggested metformin to be cardioprotective in overweight patients 413, the present data do not allow
assessment of metformin effects on endothelial function since all patients used metformin. Considering that
obesity is associated with endothelial dysfunction, and weight gain during insulin therapy with both increases
in LDL cholesterol and blood pressure it is entirely possible that improvement of endothelial function during
insulin therapy only occurs when such side effects can be avoided 414.
Insulin therapy significantly decreased serum HbA1C, FFA, and triglyceride concentrations and increased
serum free insulin concentrations, which all could have contributed to improved vascular function. Increases
in serum free insulin concentrations were due to acute or chronic effects of insulin injections. Insulin directly
attenuates vascular contraction by inhibiting voltage-dependent calcium channels 415. Acute increases in
insulin concentrations have been shown to enhance ACh-induced vasodilatation in normotensive subjects
and patients with hypertension 416. Acute hyperinsulinemia has also reversed endothelial dysfunction induced
by circulating FFAs 409. However, the insulin concentrations in these studies have been much higher than
those observed during chronic insulin therapy. Insulin therapy induced several antiatherogenic changes of
serum lipids including decreases in serum triglyceride and FFA concentrations as in previous studies 403; 407-
409. In the present study, changes in none of these parameters during insulin therapy were significantly
associated with the improvement in endothelial function. Possibly all changes contributed to this result thus
preventing any single parameter to become significant in a relatively small group of patients. Other factors
which might have contributed to enhanced endothelial function but were not measured in this study include
increased bioavailability of both endogenous and exogenous NO due to decreases in oxidative stress 417,
advanced glycosylation end products 405, small dense LDL particles 418, and susceptibility of circulating LDL to
oxidation 419. Ciprofibrate therapy, by attenuating postprandial lipemia and modifying an atherogenic
lipoprotein profile, have recently been shown to improve endothelial function both under basal and
postprandial conditions and to attenuate postprandial oxidative stress in type 2 diabetes 211. Infusions of
vitamin C have also been shown to improve endothelium-dependent vasodilation in patients with type 2
diabetes 176, but long-term antioxidant therapies have not consistently improved endothelial function in type 2
diabetes 180; 183; 184.
Insulin therapy improves insulin sensitivity of glucose uptake in type 2 diabetic patients 420-422. Insulin acts
also through other organs and tissues like liver and vascular tissue. If insulin action on vascular tissue would
also improve, it could provide a new mechanism for improved vascular function after insulin therapy. In line
with this, multiple defects in components of the insulin signaling pathway in vascular tissue of the insulin
resistant Zucker rat has been reported 137. Thus, primary defects in the insulin signaling pathway may not
manifest themselves only as relative metabolic defects but also as vascular defects.
Disturbances of hemostatic and fibrinolytic mechanisms are frequent in patients with type 2 diabetes 423; 424,
but these hemostatic abnormalities seem to persist despite improvement of glycemic control by insulin
therapy 425; 426. Insulin therapy also have effects on circulating adhesion molecules. We recently observed that
improvement in glycemic control during one year of insulin therapy was associated with a significant decrease
in serum soluble E-selectin concentrations in the FINFAT study 427. An almost identical decrease was
observed in the present study (data not shown). Finally, in the UKPDS the incidence of myocardial infarction
decreased by 16% when HbA1C improved by 0.9%, which was almost statistically significant 172. This result
was consistent with epidemiological evidence and documented that the drugs used to improve glycemic
control (insulin, sulphonylureas) were safe. This study left open the possibility that a greater improvement in
glycemic control might significantly reduce CV events. In view of these UKPDS data, the improvement in
endothelial function could also have been a consequence of improved glycemia per se.
50
7.2. Effects of oral and transdermal estradiol on markers of cardiovascular risk
Potentially beneficial effects
In study III, oral estradiol markedly improved endothelium-dependent and, to a lesser extent, endothelium-
independent blood flow in forearm resistance vessels. In this study, serum free estradiol concentrations were
identical in the oral and transdermal estradiol groups. Despite this, endothelium-dependent vasodilatation
increased over 100% in the oral estradiol group but remained virtually unchanged in the transdermal group. In
the oral estradiol group, the percent increases in endothelium-dependent flow above basal compared to
those measured at 0 weeks averaged 21-22% (low-high dose) at 2 weeks and 120-119% (low-high dose) at
12 weeks during ACh infusions. The corresponding percent increases during SNP infusions averaged 16-
28% (low-high dose) and 50-64% (low-high-dose). Total and free estradiol serum concentrations were
maximal already at 2 weeks. These data demonstrate that changes in vascular function during oral ERT
cannot be attributed to acute vascular effects of estradiol. This result is in keeping with data showing doubling
of postischemic vasodilatation between one and six months of oral HRT, but in this study both unopposed
and combination therapies were used and serum estradiol concentrations were not measured 287.
Regarding the mechanism underlying improved endothelial function, the antiatherogenic changes in serum
lipids and lipoproteins induced by oral but not transdermal estradiol should be considered. Lowering of LDL
cholesterol approximately 30% by statins improves endothelium-dependent vasodilatation in brachial 28-31 and
coronary 32; 33 arteries. When assessed using methods similar to those in the present study, endothelium-
dependent vasodilatation has improved by 33-62% 28; 31. Thus one would not expect a 9% decrease in LDL
cholesterol concentration by oral estradiol to increase endothelial function by over 100% as was now
observed. The improvement in endothelial function over that expected based on LDL cholesterol lowering
could be explained by other antiatherogenic changes in lipids and lipoproteins not observed in statin trials.
These include increases in HDL cholesterol 61; 428 and decreases in Lp(a) concentrations 429. Estrogens may
also act as an antioxidant and protect LDL from oxidation 322-324, a possibility not explored in the present
study. The small decrease in serum free testosterone concentrations could also have contributed to
beneficial effects of oral estradiol on endothelial function 430. Infusion of estrone causes vasodilatation 431.
Estradiol metabolites including estrone increase prostacyclin synthesis suggesting that 10-fold increase in
serum estrone concentrations in the oral estradiol group may have contributed in the estradiol-induced
vasodilatation.
In vitro studies have suggested several additional mechanisms via which estradiol could affect endothelial
function. Estrogens have beneficial effects on the renin-angiotensin system 432, exert antioxidant properties
322-324, and may act as calcium-blocking agents 433. In addition, estrogens, at least in high concentrations,
have also been suggested to directly stimulate NO production via increasing endothelial NOS expression 434;
435, although this has not been a consistent finding 436. Although these in vitro studies suggest that estradiol
has potentially beneficial direct effects on the endothelium, the present in vivo data question the clinical
relevance of such observations.
In study IV, estradiol therapy decreased basal peripheral vascular resistance and slightly diastolic blood
pressure by increasing peripheral blood flow. Acute administration of estradiol resulting in supraphysiological
estradiol concentrations has previously been shown to decrease peripheral vascular resistance via increases
in peripheral blood flow 281. Cessation of HRT has been reported to increase pulse wave velocity in the
femoral-dorsalis pedis region, a finding consistent with the idea that use of estrogen is associated with
decreased peripheral vascular resistance 305. In previous endothelial function studies, basal flow tended to
increase in two studies 286; 288 and was not reported in four studies 252; 279; 291; 437. The decrease in diastolic
pressure in the present study is in keeping with other data in which HRT had a small blood pressure lowering
effect 438-442. Sequential combined HRT delivered by both oral and transdermal routes caused significant and
similar falls in the daytime ambulatory blood pressure of normotensive postmenopausal women at 2 months
of treatment 443.
The 12-week treatment with oral estradiol increased significantly not only endothelium-dependent but also
endothelium-independent vasodilatation suggesting that estradiol therapy improves both endothelial and
smooth muscle cell function. The mechanism for these effects has not been determined but may be a direct
relaxing effect of estrogen on smooth muscle cells, possibly involving calcium antagonism 433; 444-446 or
increased bioavailability of NO possible through antioxidant properties of estrogen 322-324 .
51
In addition to effects on vascular function, oral estradiol had effects on circulating markers of coagulation and
fibrinolysis. Oral estradiol significantly lowered PAI-1 and tPA antigen concentrations and PAI-1 activity, and
increased D-dimer concentrations, changes consistent with increased fibrinolysis. Decreases in PAI-1 antigen
concentrations have previously been observed during oral CEE alone 288; 363 or combined with progestin 325,
and oral estradiol combined with progestin 254; 362; 367. Recently, impairment of fibrinolytic activity was found to
be associated with an increase in the risk of acute coronary events 447. Support for a relationship between
elevated plasma PAI-1 activity and myocardial infarction first came from a follow-up study of men who
survived a myocardial infarction before the age of 45 years 448. In this group, high PAI-1 activity independently
predicted reinfarction within three years of the primary event 448. Combination therapy with oral estradiol and
progestin have previously been shown to decrease tPA antigen concentrations 254. Plasma concentration of
tPA antigen has been shown to be independent predictor of subsequent myocardial infarction or sudden
death from coronary artery disease among nearly 3000 men and women followed for two years 449. Taken
together data on fibrinolysis seem quite consistent and indicate that oral estrogen alone or in combination
with progestin enhances fibrinolysis, an effect which can be considered beneficial 448.
Oral but not transdermal estradiol decreased serum soluble E-selectin concentrations. Serum soluble E-
selectin concentrations are elevated in disease states characterized with an increase in the risk of CVD such
as in type 1 and type 2 diabetes 450, in CHD itself 451, and in acute myocardial infarction 452. There are,
however, no epidemiological data regarding the predictive value of serum soluble E-selectin concentrations
for future vascular events. Thus, the significance of the decrease in this endothelial product in this and in a
previous study where CEE was used 288 remains uncertain but may reflect diminished endothelial activation
and therefore be beneficial.
Potentially harmful effects
In study V, with the use of specific and sensitive methods to measure the mass concentration of factor VII
and as well as its activity, we found oral estradiol to increase both of these markers of coagulation. Oral
estradiol therapy also caused prothrombin activation. These potentially harmful changes in coagulation
provide potential explanations for the increased risk of venous thromboembolism in women using oral HRT
223; 453. Hypothetically, the early increase in the risk for CHD events in HERS could have been due to an
immediate prothrombotic effect of HRT that might be gradually outweighed by beneficial effects on
endothelial function 223.
Oral estradiol increased serum CRP concentrations, as was observed during combination therapy with oral
CEE and different progestins in the PEPI-trial 345. Although the concentration of CRP was not significantly
higher in oral estradiol users than that of the women on placebo, the changes in CRP were significantly
different. Increased concentrations of CRP are associated with increased CV risk among otherwise healthy
women 354. Also, individuals with pre-existing CHD 348, and those at greatest risk of future CV events 349; 350
have increased serum concentrations of CRP. Thus, the present data suggest that estradiol may have
adverse effects on CV risk through an increase in inflammation although the origin of CRP cannot be
determined based on the present study. It is now clear that CRP originates from the vessel wall 454 as well as
from the liver 455. Considering that oral estradiol has profound stimulatory effects on multiple hepatic proteins
such as SHBG in the present study, it would seem likely that the increased circulating CRP concentrations
reflected hepatic synthesis. Such an increase would be predicted to be inconsequential for vascular function.
CV risk markers not altered by estrogen replacement therapy
Estradiol therapy had no effect on basal arterial stiffness, but decreased peripheral vascular resistance by
increasing blood flow. The finding of no effect on basal arterial stiffness is in keeping with many 300-302 but not
all 303; 304 cross-sectional studies. Regarding intervention studies, Waddell et al. 305 measured systemic
arterial compliance and pulse wave velocity in postmenopausal women on and 4 weeks off HRT. Large artery
compliance remained unchanged, while pulse wave velocity in the femoral-dorsalis pedis region increased
significantly after withdrawal of HRT suggesting that hormonal modulation of distal arterial vascular tone may
account for changes in arterial compliance 305. Short-term (up to 6 months) oral 306 and transdermal 307 HRT
have been shown to reduce the pulsatility index , which reflects impedance to blood flow distal to the point of
sampling, in carotid artery. HRT has also been associated with a reduction in pulsatility index in peripheral
arteries 308. All of these changes could represent a direct or indirect effect of estrogen on peripheral
resistance arteries.
Estradiol had no effect on insulin action on glucose metabolism. This negative finding adds to the list of other
negative studies which used either 100 µg transdermal estradiol or 1.25 mg oral CEE 330, 50 µg transdermal
52
estradiol and oral norethisterone 329 or continuous combined oral estradiol-norethisterone acetate 331. In view
of these negative data, it is of interest that women, when compared to equally fit men, are more insulin
sensitive and utilize greater amounts of glucose per kg muscle tissue, although women always have a greater
percentage fat of their body weight than men 456; 457. In men and postmenopausal women, serum estradiol
concentrations are below 100 pmol/l while they average 200-300 pmol/l during the follicular and 500-600
pmol/l during the luteal phase in premenopausal women. In the present study, estradiol therapy increased
serum estradiol concentrations to those characterizing premenopausal women during the follicular phase.
Despite of this, insulin sensitivity remained unchanged suggesting that the gender difference in insulin
sensitivity cannot be explained by differences in serum estradiol concentrations between men and women.
Regarding the effects of androgens on insulin sensitivity, testosterone administration to castrated male rats
has been shown to decrease insulin sensitivity 458. Short-term (10-12 days) methyltestosterone administration
to normally menstruating women has also been shown to cause a significant reduction in whole body insulin
sensitivity 459. These findings are also consistent with induction of insulin resistance by 4 months of
intramuscular testosterone injections in women 460. These data raise the possibility that androgens rather
than estrogens regulate insulin sensitivity of glucose uptake.
7.3. Concluding remarks
Atherosclerosis kills more people than any other disease in the westernized world. The present studies
emphasize the importance of assessing effects of various treatments on CV risk markers and early
alterations in vascular function in addition to studying their effects on symptoms (i.e. effects of insulin therapy
on glycemia or ERT on postmenopausal symptoms.) Results from study I adds to the growing evidence that
even mild increases in glucose concentrations may be associated with, or serve as markers of, endothelial
dysfunction and more importantly increased risk of CV disease. The finding of endothelial dysfunction in vivo
in individuals with IFG supports use of fasting glucose measurements in the identification of individuals who
should be targets for intensified lifestyle and possibly pharmaceutical interventions to prevent CV disease. It
is presently unclear whether endothelial dysfunction precedes IFG and perhaps insulin resistance or is a
consequence of some other metabolic abnormality. It is of some interest in this respect that statins, which
effectively enhance endothelial function 28-33, but have no effects on insulin sensitivity, prevent the
development of type 2 diabetes 461.
The frequency of coronary heart disease, stroke and peripheral vascular disease are all several-fold higher in
patients with type 2 diabetes than in non-diabetic subjects. It is therefore important to consider effects of
various treatments not only on glycemia but on various markers of CV risk and on direct measures of
vascular function in this high-risk group. In the present study combination therapy with bedtime insulin
ameliorated several metabolic abnormalities including hyperglycemia, hypertriglyceridemia, and high FFA
concentrations without inducing weight gain, increases in LDL cholesterol or blood pressure. These treatment
effects were associated with significant improvements of both endothelium-dependent and -independent
vasodilatation in forearm resistance vessels. These changes can be considered potentially antiatherogenic
and support the view that insulin therapy, either directly or indirectly, has beneficial rather than harmful effects
on vascular function and CV risk factors.
Many observational studies have found a lower rate of coronary heart disease in women who use
postmenopausal HRT than in non-users. However, the observed association between HRT and reduced risk
of CHD might be attributable to selection biases or other problems of observational studies. We determined
in a randomized trial effects of oral and transdermal ERT on various CV risk markers and on endothelial
function in healthy postmenopausal women. Although both oral and transdermal estradiol relieved
postmenopausal symptoms, effects of oral and transdermal estradiol on CV risk markers were strikingly
different. Oral but not transdermal estradiol markedly improved endothelium-dependent and to a lesser extent
endothelium-independent vasodilatation in forearm resistance vessels and induced several potentially
antiatherogenic changes in serum lipids and lipoproteins. Oral but not transdermal estradiol also increased
markers of fibrinolytic activity and decreased soluble E-selectin concentrations. In contrast to these beneficial
effects oral but not transdermal estradiol induced potentially harmful changes in coagulation and increased
serum CRP concentrations. These differential effects could not be explained by serum free estradiol
concentrations which were identical in the oral and transdermal estradiol groups. Oral estradiol induced
changes in serum lipids and lipoproteins may contribute but most likely do not fully explain these differential
effects of oral and transdermal estradiol on endothelial function. Thus, our data suggest that oral rather than
transdermal estradiol should be chosen as the estradiol component of HRT in healthy postmenopausal
women.
53
8. Summary
The results of studies I-V can be summarized as follows:
1) Impaired fasting glucose is associated with impaired endothelium-dependent vasodilatation in vivo
(Study I).
2) Endothelial dysfunction characterizes patients with type 2 diabetes. Successful insulin therapy normalizes
both endothelium-dependent and -independent vasodilatation in patients with type 2 diabetes (Study II)
3) Oral and transdermal estradiol have different effects on endothelium-dependent vasodilatation. Oral
estradiol markedly improves endothelium-dependent and, to a lesser extent, endothelium-independent blood
flow in forearm resistance vessels. In contrast, transdermal estradiol has no significant effects on
endothelium-dependent vasodilatation. (Study III)
4) Neither oral nor transdermal estradiol changes insulin actions on glucose metabolism, peripheral blood
flow or arterial stiffness. (Study IV)
5) Oral estradiol increases markers of fibrinolytic activity, decreases serum soluble E-selectin concentrations
and induces potentially antiatherogenic changes in serum lipids and lipoproteins. In contrast to these
beneficial effects, oral estradiol changes markers of coagulation towards hypercoagulability, and increases
serum CRP concentrations. Transdermal estradiol has no effects on any of these parameters. (Study V)
54
9. Acknowledgements
The work for this thesis was carried out at the Department of Medicine, Division of Diabetes at the Helsinki
University Central Hospital, during the years 1997-2000. I am deeply grateful to Professor Marja-Riitta
Taskinen for her support, pleasant collaboration, and for placing the research facilities at my disposal.
I want to express my deepest gratitude to my supervisor Professor Hannele Yki-Järvinen for her expert
guidance to the world of science. Her unique knowledge, brilliant ideas and enthusiasm have been invaluable
in this work and her help cannot be concretely measured but is sincerely appreciated.
I am thankful to Professor John Cockcroft and Professor Olavi Ylikorkala for constructive review of this
thesis.
The colleagues I worked with made the hours spent at work very pleasant and inspiring and have become
good friends. I especially want to thank Jukka Westerbacka, Robert Bergholm, and Antti Virkamäki who all
have been extremely helpful to me in numerous ways, and whose friendship, supportive discussions and
positive attitude have always brought my spirits up, and, Kati Ylitalo, Leena Ryysy, Juha Vakkilainen, Anna
Schlenzka, Elena Korsheninnikova, and Anneli Seppälä-Lindroos, for pleasant co-operation and especially for
your friendship. I owe my sincere thanks to my collaborators Outi Hovatta, Tiina Hakala-Ala-Pietilä, Anders
Hamsten, Angela Silveira, Sari Mäkimattila, Per-Henrik Groop, and Christian Ehnholm for pleasant and
productive co-operation. It has been a great pleasure to me to have new fellows Marjo Tamminen, Mirja
Tiikkainen, and Jussi Sutinen in our research group, and I thank them for their support in the last phases of
this work. I also want to express my thanks to Sari Haapanen, Kati Tuomola, Katja Tuominen, Ulla
Grundstedt, Hannele Hilden, Helinä Perttunen-Nio, and Ritva Marjanen for excellent technical assistance. I
owe my thanks to Maaria Puupponen for skillful secretarial and other kind of help, and Soile Aarnio for
drawing figures for all kind of purposes - conferences, papers and this thesis.
I am grateful to Tatu Miettinen for inviting me to join the Wednesday jogging group. I can only wonder if he
really knew all the consequences of that invitation. Running and especially discussions before, during and
after this run with Sami Mustajoki, Jaakko Kaukonen, Heikki Relas, Nina Lindbohm, Kristian Paavonen,
Mikael Fraunberg, and all the other occasional visitors gave me enormous amount of energy and is highly
appreciated. I owe my thanks to Pauliina Peltoniemi from Turku and Hanna-Maaria Lakka from Kuopio for
excellent congress company, and for Laura Oksanen and Alpo Vuorio for help and discussions while
preparing this thesis.
These studies could not have been performed without the subjects and patients who volunteered to
participate. My greatest gratitude for their help.
I want to thank all my friends for being my friends especially Reita Nyberg and Topi Miettinen, Saku and
Hanna Kotola, Tommi Vehkavaara and Satu Hintikka for also making us godparents, and my parents-in-law
Elina Laaksi and Olli Vehkavaara for all the support and care.
My warmest thanks belong to my dear parents Eila and Pertti Raimiala. They have given me genes loaded
with energy and positive attitude towards life, accompanied with their endless love and support. I also want to
thank my dear twin brother Tero for sharing life with me since we were born, and my godparents Tapio and
Kaija Nieminen for introducing me the world of medicine and for support and encouragement during these
years.
Last, but definitely not least, I want to thank my dear husband Mikko for his support, and for his enormous
patience when his newly-wed wife left time after time abroad for several weeks. His love has been essential
for my work.
The financial support from the Academy of Finland is warmly acknowledged. This work was also financially
supported by grants from the Foundation for Diabetes Research and the Urho Känkäre Foundation.
Helsinki, May 2001
55
References
1. Lip GY: Target organ damage and the prothrombotic state in hypertension. Hypertension
2000;36:975-977.
2. Furchgott RF, Zawadzki JV: The obligatory role of endothelial cells in the relaxation of arterial smooth
muscle by acetylcholine. Nature 1980;288:373-376.
3. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ: Compensatory enlargement of
human atherosclerotic coronary arteries. N Engl J Med 1987;316:1371-1375.
4. Vogel RA: Cholesterol lowering and endothelial function. Am J Med 1999;107:479-487.
5. Gimbrone MAJ, Bevilacqua MP, Cybulsky MI: Endothelial-dependent mechanisms of leukocyte
adhesion in inflammation and atherosclerosis. Ann N Y Acad Sci 1990;598:77-85.
6. Ross R: Atherosclerosis--an inflammatory disease. N Engl J Med 1999;340:115-126.
7. Libby P, Ross R, Topol EJ: Cytokines and growth regulatory molecyles., in Fuster V, Ross R, Topol
EJ (eds): Atherosclerosis and coronary artery disease. Vol. 1. Philadelphia, Lippincott-
Raven, 1996, pp 585-594.
8. Raines EW, Rosenfeld ME, Ross R: The role of macrophages., in Fuster V, Ross R, Topol EJ (eds):
Atherosclerosis and coronary artery disease. Vol. 1. Philadelphia, Lippincott-Raven, 1996, pp
539-555.
9. Falk E, Shah PK, Fuster V: Pathogenesis of plaque disruption, in Fuster V, Ross R, Topol EJ (eds):
Atherosclerosis and coronary artery disease. Vol. 2. Philadelphia, Lippincott-Raven, 1996, pp
492-510.
10. Stary HC, Chandler AB, Dinsmore RE, Fuster V, Glagov S, Insull WJ, Rosenfeld ME, Schwartz CJ,
Wagner WD, Wissler RW: A definition of advanced types of atherosclerotic lesions and a
histological classification of atherosclerosis. A report from the Committee on Vascular
Lesions of the Council on Arteriosclerosis, American Heart Association. Circulation
1995;92:1355-1374.
11. Moncada S, Gryglewski R, Bunting S, Vane JR: An enzyme isolated from arteries transforms
prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation.
Nature 1976;263:663-665.
12. Palmer RMJ, Ashton DS, Moncada S: Vascular endothelial cells synthesize nitric oxide from L-
arginine. Nature 1988;333:664-666.
13. Marletta MA: Nitric oxide synthase structure and mechanism. J Biol Chem 1993;268:12231-12234.
14. Boulanger CM, Vanhoutte PM: G proteins and endothelium-dependent relaxations. J Vasc Res
1997;34:175-185.
15. Vane JR, Anggard EE, Botting RM: Regulatory functions of the vascular endothelium. N Engl J Med
1990;323:27-36.
16. Arnold WP, Mittal CK, Katsuki S, Murad F: Nitric oxide activates guanylate cyclase and increases
guanosine 3':5'-cyclic monophosphate levels in various tissue preparations. Proc Natl Acad
Sci U S A 1977;74:3203-3207.
17. Kato T, Iwama Y, Okumura K, Hashimoto H, Ito T, Satake T: Prostaglandin H2 may be the
endothelium-derived contracting factor released by acetylcholine in the aorta of the rat.
Hypertension 1990;15:475-481.
56
18. Auch-Schwelk W, Katusic ZS, Vanhoutte PM: Thromboxane A2 receptor antagonists inhibit
endothelium-dependent contractions. Hypertension 1990;15:699-703.
19. Creager MA, Cooke JP, Mendelsohn ME, Gallagher SJ, Coleman SM, Loscalzo J, Dzau VJ: Impaired
vasodilation of forearm resistance vessels in hypercholesterolemic humans. J Clin Invest
1990;86:228-234.
20. Panza JA, Quyyumi AA, Brush JE, Epstein SE: Abnormal endothelium-dependent vascular relaxation
in patients with essential hypertension. N Engl J Med 1990;323:22-27.
21. Chowienczyk PJ, Cockcroft JR, Ritter JM: Differential inhibition by NG-monomethyl-L-arginine of
vasodilator effects of acetylcholine and methacholine in human forearm vasculature. Br J
Pharmacol 1993;110:736-738.
22. Kneale BJ, Chowienczyk PJ, Cockcroft JR, Coltart DJ, Ritter JM: Vasoconstrictor sensitivity to
noradrenaline and NG-monomethyl-L-arginine in men and women. Clin Sci (Colch)
1997;93:513-518.
23. Chowienczyk PJ, Cockcroft JR, Ritter JM: Blood flow responses to intra-arterial acetylcholine in man:
effects of basal flow and conduit vessel length. Clin Sci (Colch) 1994;87:45-51.
24. Cohen RA, Vanhoutte PM: Endothelium-dependent hyperpolarization. Beyond nitric oxide and cyclic
GMP. Circulation 1995;92:3337-3349.
25. Sorensen KE, Kristensen IB, Celermajer DS: Atherosclerosis in the human brachial artery. J Am Coll
Cardiol 1997;29:318-322.
26. Neunteufl T, Katzenschlager R, Hassan A, Klaar U, Schwarzacher S, Glogar D, Bauer P, Weidinger
F: Systemic endothelial dysfunction is related to the extent and severity of coronary artery
disease. Atherosclerosis 1997;129:111-118.
27. Lieberman EH, Gerhard MD, Uehata A, Selwyn AP, Ganz P, Yeung AC, Creager MA: Flow-induced
vasodilatation of the human brachial artery is impaired in patients <40 years of age with
coronary artery disease. Am J Cardiol 1996;78:1210-1214.
28. O'Driscoll G, Green D, Taylor RR: Simvastatin, an HMG-coenzyme A reductase inhibitor, improves
endothelial function within 1 month. Circulation 1997;95:1126-1131.
29. Dupuis J, Tardif JC, Cernacek P, Theroux P: Cholesterol reduction rapidly improves endothelial
function after acute coronary syndromes. The RECIFE (reduction of cholesterol in ischemia
and function of the endothelium) trial. Circulation 1999;99:3227-3233.
30. Stroes ES, Koomans HA, de Bruin TW, Rabelink TJ: Vascular function in the forearm of
hypercholesterolaemic patients off and on lipid-lowering medication. Lancet 1995;346:467-
471.
31. John S, Schlaich M, Langenfeld M, Weihprecht H, Schmitz G, Weidinger G, Schmieder RE:
Increased bioavailability of nitric oxide after lipid-lowering therapy in hypercholesterolemic
patients: a randomized, placebo-controlled, double-blind study. Circulation 1998;98:211-216.
32. Egashira K, Hirooka Y, Kai H, Sugimachi M, Suzuki S, Inou T, Takeshita A: Reduction in serum
cholesterol with pravastatin improves endothelium-dependent coronary vasomotion in
patients with hypercholesterolemia. Circulation 1994;89:2519-2524.
33. Hamasaki S, Higano ST, Suwaidi JA, Nishimura RA, Miyauchi K, Holmes DRJ, Lerman A:
Cholesterol-lowering treatment is associated with improvement in coronary vascular
remodeling and endothelial function in patients with normal or mildly diseased coronary
arteries. Arterioscler Thromb Vasc Biol 2000;20:737-743.
57
34. Anderson TJ, Uehata A, Gerhard MD, Meredith IT, Knab S, Delagrange D, Lieberman EH, Ganz P,
Creager MA, Yeung AC: Close relation of endothelial function in the human coronary and
peripheral circulations. J Am Coll Cardiol 1995;26:1235-1241.
35. Takase B, Uehata A, Akima T, Nagai T, Nishioka T, Hamabe A, Satomura K, Ohsuzu F, Kurita A:
Endothelium-dependent flow-mediated vasodilation in coronary and brachial arteries in
suspected coronary artery disease. Am J Cardiol 1998;82:1535-1538.
36. Suwaidi JA, Hamasaki S, Higano ST, Nishimura RA, Holmes DRJ, Lerman A: Long-term follow-up of
patients with mild coronary artery disease and endothelial dysfunction. Circulation
2000;101:948-954.
37. Schachinger V, Britten MB, Zeiher AM: Prognostic impact of coronary vasodilator dysfunction on
adverse long-term outcome of coronary heart disease. Circulation 2000;101:1899-1906.
38. Celermajer DS, Sorensen KE, Gooch VM, Spiegelhalter DJ, Miller OI, Sullivan ID, Lloyd JK,
Deanfield JE: Non-invasive detection of endothelial dysfunction in children and adults at risk
of atherosclerosis. Lancet 1992;340:1111-1115.
39. Berry KL, Skyrme-Jones RA, Meredith IT: Occlusion cuff position is an important determinant of the
time course and magnitude of human brachial artery flow-mediated dilation. Clin Sci (Colch)
2000;99:261-267.
40. Hata A, Reid CL, Ogata T, Tokushima T, Gardin JM: Reproducibility of brachial artery ultrasound
measurements. Echocardiography 1999;16:367-372.
41. Chowienczyk PJ, Watts GF, Cockcroft JR, Ritter JM: Impaired endothelium-dependent vasodilatation
of forearm resistance vessels in hypercholesterolemia. Lancet 1992;340:1430-1432.
42. Casino PR, Kilcoyne CM, Quyyumi AA, Hoeg JM, Panza JA: The role of nitric oxide in endothelium-
dependent vasodilation of hypercholesterolemic patients. Circulation 1993;88:2541-2547.
43. Drexler H, Zeiher AM, Meinzer K, Just H: Correction of endothelial dysfunction in coronary
microcirculation of hypercholesterolemic patients by L-arginine. Lancet 1991;338:1546-
1550.
44. Leung WH, Lau CP, Wang CK: Beneficial effect of cholesterol-lowering therapy on the coronary
endothelium-dependent relaxation in hypercholesterolemic patients. Lancet 1993;89:2519-
2524.
45. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the
Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-1389.
46. Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, Cole TG, Brown L, Warnica JW, Arnold
JM, Wun CC, Davis BR, Braunwald E: The effect of pravastatin on coronary events after
myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent
Events Trial investigators. N Engl J Med 1996;335:1001-1009.
47. The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group: Prevention
of cardiovascular events and death with pravastatin in patients with coronary heart disease
and a broad range of initial cholesterol levels. N Engl J Med 1998;339:1349-1357.
48. Levine GN, Keaney JFJ, Vita JA: Cholesterol reduction in cardiovascular disease. Clinical benefits
and possible mechanisms. N Engl J Med 1995;332:512-521.
49. Treasure CB, Klein JL, Weintraub WS, Talley JD, Shillabower ME, Kosinski AS, Zhang J, Boccuzz
SI, Cedarholm JC, Alexander RW: Beneficial effects of cholesterol-lowering therapy on the
coronary endothelium in patients with coronary artery disease. N Engl J Med 1995;332:481-
487.
58
50. Anderson TJ, Meredith IT, Yeung AC, Frei B, Selwyn AP, Ganz P: The effect of cholesterol-lowering
and antioxidant therapy on endothelium-dependent coronary vasomotion. N Engl J Med
1995;332:488-493.
51. Glorioso N, Troffa C, Filigheddu F, Dettori F, Soro A, Parpaglia PP, Collatina S, Pahor M: Effect of
the HMG-CoA reductase inhibitors on blood pressure in patients with essential hypertension
and primary hypercholesterolemia. Hypertension 1999;34:1281-1286.
52. Vita JA, Yeung AC, Winniford M, Hodgson JM, Treasure CB, Klein JL, Werns S, Kern M, Plotkin D,
Shih WJ, Mitchel Y, Ganz P: Effect of cholesterol-lowering therapy on coronary endothelial
vasomotor function in patients with coronary artery disease. Circulation 2000;102:846-851.
53. Andrews TC, Whitney EJ, Green G, Kalenian R, Personius BE: Effect of gemfibrozil +/- niacin +/-
cholestyramine on endothelial function in patients with serum low-density lipoprotein
cholesterol levels <160 mg/dl and high-density lipoprotein cholesterol levels <40 mg/dl. Am J
Cardiol 1997;80:831-835.
54. Chowienczyk PJ, Kneale BJ, Brett SE, Paganga G, Jenkins BS, Ritter JM: Lack of effect of vitamin E
on L-arginine-responsive endothelial dysfunction in patients with mild hypercholesterolaemia
and coronary artery disease. Clin Sci (Colch) 1998;94:129-134.
55. Chowienczyk PJ, Watts GF, Cockcroft JR, Brett SE, Ritter JM: Sex differences in endothelial function
in normal and hypercholesterolaemic subjects. Lancet 1994;344:305-306.
56. Chin JH, Azhar S, Hoffman BB: Inactivation of endothelial derived relaxing factor by oxidized
lipoproteins. J Clin Invest 1992;89:10-18.
57. Galle J, Bengen J, Schollmeyer P, Wanner C: Impairment of endothelium-dependent dilation in rabbit
renal arteries by oxidized lipoprotein(a). Role of oxygen-derived radicals. Circulation
1995;92:1582-1589.
58. Anderson TJ, Meredith IT, Charbonneau F, Yeung AC, Frei B, Selwyn AP, Ganz P: Endothelium-
dependent coronary vasomotion relates to the susceptibility of LDL to oxidation in humans.
Circulation 1996;93:1647-1650.
59. Raitakari OT, Pitkänen OP, Lehtimäki T, Lahdenperä S, Iida H, Ylä-Herttuala S, Luoma J, Mattila K,
Nikkari T, Taskinen MR, Viikari JS, Knuuti J: In vivo low density lipoprotein oxidation relates
to coronary reactivity in young men. J Am Coll Cardiol 1997;30:97-102.
60. Matsuda Y, Hirata K, Inoue N, Suematsu M, Kawashima S, Akita H, Yokoyama M: High density
lipoprotein reverses inhibitory effect of oxidized low density lipoprotein on endothelium-
dependent arterial relaxation. Circ Res 1993;72:1103-1109.
61. Kuhn FE, Mohler ER, Satler LF, Reagan K, Lu DY, Rackley CE: Effects of high-density lipoprotein on
acetylcholine-induced coronary vasoreactivity. Am J Cardiol 1991;68:1425-1430.
62. Schnell GB, Robertson A, Houston D, Malley L, Anderson TJ: Impaired brachial artery endothelial
function is not predicted by elevated triglycerides. J Am Coll Cardiol 1999;33:2038-2043.
63. Chowienczyk PJ, Watts GF, Wierzbicki AS, Cockcroft JR, Brett SE, Ritter JM: Preserved endothelial
function in patients with severe hypertriglyceridemia and low functional lipoprotein lipase
activity. J Am Coll Cardiol 1997;29:964-968.
64. Vakkilainen J, Mäkimattila S, Seppälä-Lindroos A, Vehkavaara S, Lahdenperä S, Groop PH,
Taskinen MR, Yki-Järvinen H: Endothelial dysfunction in men with small LDL particles.
Circulation 2000;102:716-721.
65. Levine GN, Frei B, Koulouris SN, Gerhard MD, Keaney JFJ, Vita JA: Ascorbic acid reverses
endothelial vasomotor dysfunction in patients with coronary artery disease. Circulation
1996;93:1107-1113.
59
66. Gokce N, Keaney JFJ, Frei B, Holbrook M, Olesiak M, Zachariah BJ, Leeuwenburgh C, Heinecke
JW, Vita JA: Long-term ascorbic acid administration reverses endothelial vasomotor
dysfunction in patients with coronary artery disease. Circulation 1999;99:3234-3240.
67. Gilligan DM, Sack MN, Guetta V, Casino PR, Quyyumi AA, Rader DJ, Panza JA, Cannon RO: Effect
of antioxidant vitamins on low density lipoprotein oxidation and impaired endothelium-
dependent vasodilation in patients with hypercholesterolemia. J Am Coll Cardiol
1994;24:1611-1617.
68. Title LM, Cummings PM, Giddens K, Genest JJJ, Nassar BA: Effect of folic acid and antioxidant
vitamins on endothelial dysfunction in patients with coronary artery disease. J Am Coll
Cardiol 2000;36:758-765.
69. Garcia CE, Kilcoyne CM, Cardillo C, Cannon RO, Quyyumi AA, Panza JA: Evidence that endothelial
dysfunction in patients with hypercholesterolemia is not due to increased extracellular nitric
oxide breakdown by superoxide anions. Am J Cardiol 1995;76:1157-1161.
70. Cooke JP: Does ADMA cause endothelial dysfunction? Arterioscler Thromb Vasc Biol
2000;20:2032-2037.
71. Sackett DL, Gibson RW, Bross ID, Pickren JW: Relation between aortic atherosclerosis and the use
of cigarettes and alcohol. An autopsy study. N Engl J Med 1968;279:1413-1420.
72. Jonas MA, Oates JA, Ockene JK, Hennekens CH: Statement on smoking and cardiovascular
disease for health care professionals. American Heart Association. Circulation
1992;86:1664-1669.
73. Chalon S, Moreno HJ, Benowitz NL, Hoffman BB, Blaschke TF: Nicotine impairs endothelium-
dependent dilatation in human veins in vivo. Clin Pharmacol Ther 2000;67:391-397.
74. Church DF, Pryor WA: Free-radical chemistry of cigarette smoke and its toxicological implications.
Environ Health Perspect 1985;64:111-126.
75. Celermajer DS, Sorensen KE, Georgakopoulos D, Bull C, Thomas O, Robinson J, Deanfield JE:
Cigarette smoking is associated with dose-related and potentially reversible impairment of
endothelium-dependent dilation in healthy young adults. Circulation 1993;88:2149-2155.
76. Zeiher AM, Schachinger V, Minners J: Long-term cigarette smoking impairs endothelium-dependent
coronary arterial vasodilator function. Circulation 1995;92:1094-1100.
77. Kaufmann PA, Gnecchi-Ruscone T, di Terlizzi M, Schafers KP, Luscher TF, Camici PG: Coronary
heart disease in smokers: vitamin C restores coronary microcirculatory function. Circulation
2000;102:1233-1238.
78. Celermajer DS, Adams MR, Clarkson P, Robinson J, McCredie R, Donald A, Deanfield JE: Passive
smoking and impaired endothelium-dependent arterial dilatation in healthy young adults. N
Engl J Med 1996;334:150-154.
79. Raitakari OT, Adams MR, McCredie RJ, Griffiths KA, Celermajer DS: Arterial endothelial dysfunction
related to passive smoking is potentially reversible in healthy young adults. Ann Intern Med
1999;130:578-581.
80. Heitzer T, Just H, Munzel T: Antioxidant vitamin C improves endothelial dysfunction in chronic
smokers. Circulation 1996;94:6-9.
81. Natali A, Sironi AM, Toschi E, Camastra S, Sanna G, Perissinotto A, Taddei S, Ferrannini E: Effect of
vitamin C on forearm blood flow and glucose metabolism in essential hypertension.
Arterioscler Thromb Vasc Biol 2000;20:2401-2406.
60
82. Raitakari OT, Adams MR, McCredie RJ, Griffiths KA, Stocker R, Celermajer DS: Oral vitamin C and
endothelial function in smokers: short-term improvement, but no sustained beneficial effect.
J Am Coll Cardiol 2000;35:1616-1621.
83. Neunteufl T, Priglinger U, Heher S, Zehetgruber M, Soregi G, Lehr S, Huber K, Maurer G, Weidinger
F, Kostner K: Effects of vitamin E on chronic and acute endothelial dysfunction in smokers. J
Am Coll Cardiol 2000;35:277-283.
84. Heitzer T, Yla HS, Wild E, Luoma J, Drexler H: Effect of vitamin E on endothelial vasodilator function
in patients with hypercholesterolemia, chronic smoking or both. J Am Coll Cardiol
1999;33:499-505.
85. Collins R, Peto R, MacMahon S, Hebert P, Fiebach NH, Eberlein KA, Godwin J, Qizilbash N, Taylor
JO, Hennekens CH: Blood pressure, stroke, and coronary heart disease. Part 2, Short-term
reductions in blood pressure: overview of randomised drug trials in their epidemiological
context. Lancet 1990;335:827-838.
86. Panza JA, Quyyumi AA, Callahan TS, Epstein SE: Effect of antihypertensive treatment on
endothelium-dependent vascular relaxation in patients with essential hypertension. J Am
Coll Cardiol 1993;21:1145-1151.
87. Linder L, Kiowski W, Buhler FR, Luscher TF: Indirect evidence for release of endothelium-derived
relaxing factor in human forearm circulation in vivo. Blunted response in essential
hypertension. Circulation 1990;81:1762-1767.
88. Cockcroft JR, Chowienczyk PJ, Benjamin N, Ritter JM: Preserved endothelium-dependent
vasodilatation in patients with essential hypertension. N Engl J Med 1994;330:1036-1040.
89. Higashi Y, Sasaki S, Kurisu S, Yoshimizu A, Sasaki N, Matsuura H, Kajiyama G, Oshima T: Regular
aerobic exercise augments endothelium-dependent vascular relaxation in normotensive as
well as hypertensive subjects: role of endothelium-derived nitric oxide. Circulation
1999;100:1194-1202.
90. Schiffrin EL, Park JB, Intengan HD, Touyz RM: Correction of arterial structure and endothelial
dysfunction in human essential hypertension by the angiotensin receptor antagonist losartan.
Circulation 2000;101:1653-1659.
91. Higashi Y, Sasaki S, Nakagawa K, Ueda T, Yoshimizu A, Kurisu S, Matsuura H, Kajiyama G, Oshima
T: A comparison of angiotensin-converting enzyme inhibitors, calcium antagonists, beta-
blockers and diuretic agents on reactive hyperemia in patients with essential hypertension: a
multicenter study. J Am Coll Cardiol 2000;35:284-291.
92. Rizzoni D, Muiesan ML, Porteri E, Castellano M, Zulli R, Bettoni G, Salvetti M, Monteduro C, Agabiti-
Rosei E: Effects of long-term antihypertensive treatment with lisinopril on resistance arteries
in hypertensive patients with left ventricular hypertrophy. J Hypertens 1997;15:197-204.
93. Creager MA, Roddy MA: Effect of captopril and enalapril on endothelial function in hypertensive
patients. Hypertension 1994;24:499-505.
94. Muiesan ML, Salvetti M, Monteduro C, Rizzoni D, Zulli R, Corbellini C, Brun C, Agabiti-Rosei E: Effect
of treatment on flow-dependent vasodilation of the brachial artery in essential hypertension.
Hypertension 1999;33:575-580.
95. Kiowski W, Erne P, Linder L, Buhler FR: Arterial vasodilator effects of the dihydropyridine calcium
antagonist amlodipine alone and in combination with verapamil in systemic hypertension.
Am J Cardiol 1990;66:1469-1472.
96. Ludmer PL, Selwyn AP, Shook TL, Wayne RR, Mudge GH, Alexander RW, Ganz P: Paradoxical
vasoconstriction induced by acetylcholine in atherosclerotic coronary arteries. N Engl J Med
1986;315:1046-1051.
61
97. Cox DA, Vita JA, Treasure CB, Fish RD, Alexander RW, Ganz P, Selwyn AP: Atherosclerosis impairs
flow-mediated dilation of coronary arteries in humans. Circulation 1989;80:458-465.
98. Zeiher AM, Drexler H, Wollschlager H, Just H: Modulation of coronary vasomotor tone in humans.
Progressive endothelial dysfunction with different early stages of coronary atherosclerosis.
Circulation 1991;83:391-401.
99. Vita JA, Treasure CB, Nabel EG, McLenachan JM, Fish RD, Yeung AC, Vekshtein VI, Selwyn AP,
Ganz P: Coronary vasomotor response to acetylcholine relates to risk factors for coronary
artery disease. Circulation 1990;81:491-497.
100. Nishimura RA, Lerman A, Chesebro JH, Ilstrup DM, Hodge DO, Higano ST, Holmes DRJ, Tajik AJ:
Epicardial vasomotor responses to acetylcholine are not predicted by coronary
atherosclerosis as assessed by intracoronary ultrasound. J Am Coll Cardiol 1995;26:41-49.
101. Vogel RA, Corretti MC: Estrogens, progestins, and heart disease: can endothelial function divine the
benefit? Circulation 1998;97:1223-1226.
102. Glasser SP, Arnett DK, McVeigh GE, Finkelstein SM, Bank AJ, Morgan DJ, Cohn JN: Vascular
compliance and cardiovascular disease: a risk factor or a marker? Am J Hypertens
1997;10:1175-1189.
103. Kelly R, Hayward C, Avolio A, O'Rourke M: Noninvasive determination of age-related changes in the
human arterial pulse. Circulation 1989;80:1652-1659.
104. Stefanadis C, Stratos C, Boudoulas H, Kourouklis C, Toutouzas P: Distensibility of the ascending
aorta: comparison of invasive and non-invasive techniques in healthy men and in men with
coronary artery disease. Eur Heart J 1990;11:990-996.
105. Dart A, Silagy C, Dewar E, Jennings G, McNeil J: Aortic distensibility and left ventricular structure and
function in isolated systolic hypertension. Eur Heart J 1993;14:1465-1470.
106. Pitsavos C, Toutouzas K, Dernellis J, Skoumas J, Skoumbourdis E, Stefanadis C, Toutouzas P:
Aortic stiffness in young patients with heterozygous familial hypercholesterolemia. Am Heart
J 1998;135:604-608.
107. O'Rourke MF, Mancia G: Arterial stiffness. J Hypertens 1999;17:1-4.
108. Wilkinson IB, Fuchs SA, Jansen IM, Spratt JC, Murray GD, Cockcroft JR, Webb DJ: Reproducibility
of pulse wave velocity and augmentation index measured by pulse wave analysis. J
Hypertens 1998;16:2079-2084.
109. Filipovsky J, Svobodova V, Pecen L: Reproducibility of radial pulse wave analysis in healthy subjects.
J Hypertens 2000;18:1033-1040.
110. Westerbacka J, Wilkinson I, Cockcroft J, Utriainen T, Vehkavaara S, Yki-Järvinen H: Diminished
wave reflection in the aorta. A novel physiological action of insulin on large blood vessels.
Hypertension 1999;33:1118-1122.
111. Wilkinson IB, MacCallum H, Flint L, Cockcroft JR, Newby DE, Webb DJ: The influence of heart rate
on augmentation index and central arterial pressure in humans. J Physiol (Lond) 2000;525
Pt 1:263-270.
112. McVeigh GE, Bratteli CW, Morgan DJ, Alinder CM, Glasser SP, Finkelstein SM, Cohn JN: Age-
related abnormalities in arterial compliance identified by pressure pulse contour analysis:
aging and arterial compliance. Hypertension 1999;33:1392-1398.
113. O'Rourke MF, Blazek JV, Morreels CL, Krovetz LJ: Pressure wave transmission along the human
aorta. Changes with age and in arterial degenerative disease. Circ Res 1968;23:567-579.
62
114. Robert L: Aging of the vascular wall and atherogenesis: role of the elastin-laminin receptor.
Atherosclerosis 1996;123:169-179.
115. Franklin SS, Larson MG, Khan SA, Wong ND, Leip EP, Kannel WB, Levy D: Does the relation of
blood pressure to coronary heart disease risk change with aging? : The Framingham Heart
study. Circulation 2001;103:1245-1249.
116. Franklin SS, Khan SA, Wong ND, Larson MG, Levy D: Is pulse pressure useful in predicting risk for
coronary heart disease? The Framingham heart study. Circulation 1999;100:354-360.
117. Vaitkevicius PV, Fleg JL, Engel JH, O'Connor FC, Wright JG, Lakatta LE, Yin FC, Lakatta EG:
Effects of age and aerobic capacity on arterial stiffness in healthy adults. Circulation
1993;88:1456-1462.
118. Sonesson B, Hansen F, Stale H, Lanne T: Compliance and diameter in the human abdominal aorta-
the influence of age and sex. Eur J Vasc Surg 1993;7:690-697.
119. Laogun AA, Gosling RG: In vivo arterial compliance in man. Clin Phys Physiol Meas 1982;3:201-
212.
120. Jonason T, Henrikssen E, Kangro T, Nilsson H, Vessby B, Ringqvist I: Stiffness of the common
carotid artery in healthy 50-year-old subjects. Clin Physiol 1997;17:569-577.
121. Salomaa V, Riley W, Kark JD, Nardo C, Folsom AR: Non-insulin-dependent diabetes mellitus and
fasting glucose and insulin concentrations are associated with arterial stiffness indexes. The
Atherosclerosis Risk in Communities (ARIC) Study. Circulation 1995;91:1432-1443.
122. Cameron JD, Dart AM: Exercise training increases total systemic arterial compliance in humans. Am
J Physiol 1994;266:H693-H701.
123. Stefanadis C, Vlachopoulos C, Tsiamis E, Diamantopoulos L, Toutouzas K, Giatrakos N, Vaina S,
Tsekoura D, Toutouzas P: Unfavorable effects of passive smoking on aortic function in men.
Ann Intern Med 1998;128:426-434.
124. McVeigh GE, Morgan DJ, Finkelstein SM, Lemay LA, Cohn JN: Vascular abnormalities associated
with long-term cigarette smoking identified by arterial waveform analysis. Am J Med
1997;102:227-231.