clinical manifestations of coronary heart disease and - diva portal

Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Medicine 1354

Clinical Manifestations ofCoronary Heart Disease and

the Metabolic SyndromeA Population-based Study in Middle-aged Men in Uppsala

BY

KRISTINA DUNDER

ACTA UNIVERSITATIS UPSALIENSISUPPSALA 2004

List of Papers

I. Dunder K, Lind L, Zethelius B, Berglund L, Lithell H. Increase in blood glucose concentration during antihypertensive treatment as a predictor of myocardial infarction: population based cohort study. BMJ 2003 Mar 29;326(7391):681-3*

II. Dunder K, Lind L, Lagerqvist B, Zethelius B, Vessby B, Lithell H. Car-diovascular risk factors for stable angina pectoris versus unheralded myocar-dial infarction. Am Heart J. 2004 Mar;147(3):502-8.**

III. Dunder K, Lind L, Zethelius B, Lithell H. A new Q/QS pattern on the resting electrocardiogram is associated with impaired insulin secretion and a poor prognosis in elderly men independently of history of myocardial infarc-tion. J Intern Med. 2004 Feb;255(2):221-8. ***

IV. Dunder K, Lind L, Zethelius B, Berglund L, Lithell H. Evaluation of a scoring scheme including proinsulin and the apolipoprotein B/apolipoprotein A1 ratio for the risk of acute coronary events in middle-aged men. Am Heart J. in press 2004

* © BMJ Publishing group ** © Elsevier Verlag *** © Blackwell Science

iv

Contents

INTRODUCTION ..........................................................................................7Coronary heart disease ...............................................................................7

Clinical presentations.............................................................................7Pathophysiology ....................................................................................8

The Metabolic syndrome............................................................................9Definitions .............................................................................................9Insulin resistance .................................................................................12Etiological factors and pathophysiology..............................................12

Coronary heart disease and the metabolic syndrome ...............................14Impaired glucose regulation ................................................................14Dyslipidemia........................................................................................15Hypertension........................................................................................16

Risk scores for coronary heart disease .....................................................19

AIMS ............................................................................................................20

MATERIALS AND METHODS..................................................................21Subjects ....................................................................................................21

The ULSAM study ..............................................................................21Study populations ................................................................................23

Data Collection.........................................................................................26Investigations at age 50 .......................................................................26Investigations at age 60 .......................................................................27Investigations at age 70 .......................................................................28Registry data ........................................................................................31Definitions ...........................................................................................31

Statistical analyses....................................................................................32General analysis...................................................................................32Specific analyses..................................................................................32

Discussion of Methods .............................................................................34Selection bias; general aspects.............................................................34Selection bias; specific aspects ............................................................35The predictive power of CHD risk factors during long follow-ups.....36

v

RESULTS AND DISCUSSION ...................................................................37Paper I ......................................................................................................37

Results .................................................................................................37Discussion............................................................................................40

Paper II .....................................................................................................41Results .................................................................................................41Discussion............................................................................................42

PAPER III ................................................................................................43Results .................................................................................................43Discussion............................................................................................47

Paper IV ...................................................................................................48Results .................................................................................................48Discussion............................................................................................49

GENERAL DISCUSSION ...........................................................................51Metabolic consequences of antihypertensive treatment ...........................51Differences in clinical manifestations of coronary heart disease .............53The metabolic syndrome and coronary heart disease; future perspectives and clinical implications...........................................................................55

CONCLUSIONS ..........................................................................................57

ACKNOWLEDGEMENTS..........................................................................58

REFERENCES .............................................................................................60

vi

Abbreviations

AUC area under the curve ACE angiotensin converting enzyme ANOVA analysis of variance BMI body mass index CABG coronary artery by pass grafting CRP C-reactive protein CVD cardiovascular disease DBP diastolic blood pressure FPG fasting plasma glucose HDL high-density lipoprotein ICD international classification of diseases LDL low density lipoprotein PAI-1 plasminogen activator inhibitor 1 PG postchallenge glucose PTCA percutanous transluminal coronary angioplasty SBP systolic blood pressure Tg triglycerides VLDL very low density lipoprotein WHO world health organisation WHR waist hip ratio ULSAM Uppsala longitudinal study of adult men

7

INTRODUCTION

Epidemiology is the quantitative study of the distribution and determinants of disease in human populations. The term epidemiology, if taken literally, means “the study of that which befalls man” (epi=”befalls, upon”, demo=”man, people”, ology=”the study of”).

The purposes and uses of epidemiology are to explain the etiology of dis-eases and conditions, to determine if epidemiological data are consistent with current scientific knowledge and hypotheses, for planning and evalua-tion of public health measures and to describe the natural history of dis-eases1. However, absolute proof of causality can only come from interven-tion studies.

The studies in this thesis are epidemiological in their character, and exam-ine the relationships between different aspects of coronary heart disease (CHD) and the metabolic syndrome.

Coronary heart disease During the past decades a great deal of knowledge concerning the patho-physiology of CHD has been achieved, and hypertension, smoking, hyperlip-idemia and diabetes are some of the abnormalities that are generally ac-cepted as risk factors2. However, despite identification of important risk factors CHD remains the leading cause of death in Europe3 and in the United States4 and by 2010 cardiovascular disease is estimated to be the leading cause of death also in the developing countries (WHO).

Clinical presentations The clinical presentation of CHD can vary substantially from acute myocar-dial infarction (MI) without previous signs of coronary ischemia, to stable angina pectoris that may exist for many years without acute events. When studying coronary angiograms, it has been shown that patients with chronic stable angina generally have more severe stenotic lesions than those with acute unheralded MI 5-7, suggesting that these two presentations of coronary atherosclerosis, at least in part, may develop in different ways.

During the past decade some studies have tried to answer the question whether the differences in clinical presentation are due to different risk fac-

8

tor profiles, but no consistent pattern has emerged8-10. Therefore further studies concerning the different clinical presentations of CHD are needed.

It is not unusual to find Q/QS-patterns on the resting electrocardiogram (ECG) at a standard routine examination in subjects with a normal ECG at the previous examination. In some subjects it is clear that a MI has occurred, but in many other cases the MI is unrecognised, i.e. no history of typical symptoms suggesting a MI is revealed.

In the case of unrecognised MI, the risk factor profile and prognosis ap-pear to be almost the same as for recognised MI11-16. Some studies have indicated an increased prevalence of diabetes mellitus and hypertension in subjects with unrecognised MI15 17 18 and it has been proposed that auto-nomic neuropathy due to diabetes may explain this association 19 20. How-ever, others have failed to prove that diabetics are more prone to unrecog-nised MI than non-diabetics21 22.

The prognostic significance of different ECG abnormalities, such as Q-wave or QS-pattern, has been extensively examined in several epidemiologi-cal studies and a strong association with future cardiovascular morbidity and mortality has been found23-32. Several of the studies have adjusted the results for known cardiovascular risk factors, but few have investigated whether the predictive power of the Q/QS-pattern is dependent of a history of MI or not.

PathophysiologyThe common pathophysiological background to CHD is coronary atheroscle-rosis with subsequent plaque formation33-35. The initial step of atherosclero-sis is migration of lipids and inflammatory cells into the intima of the coro-nary arteries. This leads to a thickening of the intima that subsequently pro-gresses to a plaque with a lipid core and a fibrous cap. These plaques pro-gress slowly and due to a remodelling of the vessel wall leading to an increased diameter36, the lumen of the vessel can be maintained during sev-eral years and the patient may be asymptomatic.

The major pathophysiological difference between acute coronary syn-dromes and stable angina pectoris is rupture of the atherosclerotic plaque with subsequent thrombosis formation that causes the acute events. Plaque rupture occurs independently of lesion size and degree of stenosis37, and some plaques seem to be more vulnerable than others are. The vulnerable plaque is characterized by a large lipid core, a high content of macrophages and a thin fibrous cap38.

Both external forces, such as circumferential wall stress and blood flow characteristics and internal forces may be involved in plaque disruption. Inflammatory cell activity probably has an important role in plaque disrup-tion, as well as in the pathology of atherosclerosis. The role of the macro-phages has been studied more in detail, but also T-lymphocytes seem to be involved in the development of atherosclerosis39. Macrophages can secrete

9

metalloproteinases that can degrade proteins in the fibrous cap and thus make it more prone to rupture40 41. Accordingly, CRP, a sensitive marker of inflammation, has been proven to be a significant predictor of CHD in sev-eral studies42 43.

The Metabolic syndrome Definitions In 1988 Reaven suggested the existence of a syndrome, syndrome X, includ-ing a number of metabolic disorders such as hypertension, hypertriglyc-eridemia, low HDL-cholesterol, and glucose intolerance. Insulin resistance was suggested as the common antecedent behind these aberrations44, and therefore the syndrome is sometimes called the insulin resistance syndrome. However, the WHO has chosen the term metabolic syndrome45, since it is considered that current data cannot establish insulin resistance as the cause of all the components.

Over the years several definitions and explanations of the syndrome have emerged, and several other metabolic disturbances such as abdominal obe-sity, microalbuminuria, chronic subclinical inflammation, impaired fibri-nolysis and left ventricular hypertrophy have been suggested to be parts of the syndrome.

The WHO definition45 is to be applied to both diabetic and non-diabetic subjects, while the EGIR (the European Group for the Study of Insulin Re-sistance) definition46 is defined only for a non-diabetic population. A more recent definition has been presented in the Summary of the Third Report of the National Cholesterol Education Program (NCEP)47 (Table 1).

The prevalence of the syndrome depends on the definition used. In an in-vestigation of eight European studies the frequency of the syndrome in-creased with age and was more common in men than women. In non-diabetic subjects the frequency of the WHO syndrome was 7-36% for men and 5-22% for women in the ages 40-55 years, and was prevalent in 1-22% of men and 1-14% of women using the EGIR definition48. In the United States the prevalence is estimated to 24%, increasing to 44% in adults 60 years of age49.

10

Table 1. Definitions of the metabolic syndrome

WHO

Risk Factor Defining level

Insulin resistance Glucose uptake below lowest quartile for back-ground population

Impaired glucose regu-lation

FPG 6.1 mmol/l, 5.6 for venous or capillary whole blood, and/or 2-hour PG 7.8, 6.7 for venous or capillary whole blood

Hypertension SBP 140 mmHg and/or DBP 90 mmHg

Dyslipidemia S-Tg 1.7 mmol/l and/or HDL cholesterol<0.9 ( ), <1.0( )

Central obesity WHR>0.90( ), >0.85( ) and/or BMI>30 kg/m2

Microalbuminuria Urinary albumin 20µg/min

(At least one of insulin resistance and impaired glucose regulation, and two more of the other components)

11

EGIR


Hyperinsulinemia Fasting insulin concentration above the upper quartile for the non-diabetic subjects

Hyperglycemia FPG 6.1mmol/l, 5.6mmol/l for venous or cap-illary whole blood

Hypertension SBP 140mmHg and/or DBP 90 mmHg or treatment for hypertension

Dyslipidemia S-Tg 2.0 mmol/l and/or HDL cholesterol<1.0 mmol/l or treatment for hyperlipidemia

Central obesity Waist circumference 94cm ( ), 80cm ( )

(Hyperinsulinemia and two more of the other components)

NCEP


Hyperglycemia FPG 6.1 mmol/l

Hypertension 130/ 85 mmHg

High triglycerides 1.7 mmol/l

Low HDL cholesterol <1.0 mmol/l ( ), <1.2 mmol/l ( )

Abdominal obesity Waist circumference>102 cm ( ), >88 cm ( )

(Three or more components)

12

Insulin resistanceInsulin sensitivity is a term generally used to define the ability of insulin to mediate glucose disposal in skeletal muscle, but it also includes insulin’s ability to suppress lipolysis in the adipose tissue50 and gluconeogenesis in the liver51.

Reduced insulin sensitivity, denoted insulin resistance, is common in type 2 diabetes mellitus, but can also be found in non-diabetic subjects with nor-mal glucose tolerance52. In normoglycemic subjects a compensatory in-creased pancreatic insulin secretion can maintain the fasting blood glucose at a normal level despite of prevailing insulin resistance. However, most often the ability of the beta cells to produce enough insulin subsequently declines over time in insulin resistant subjects. This leads to lower glucose uptake in skeletal muscle, increased levels of free fatty acids, less inhibition of hepatic glucose production, and thus subsequent hyperglycemia50.

The gold standard for measuring insulin sensitivity in skeletal muscle is the hyperinsulineamic euglycemic clamp method53. Since this is a rather complicated method not used in clinical practice, fasting serum insulin has often been used as a proxy of insulin sensitivity in epidemiological studies. Elevated concentrations of proinsulin have been associated with measure-ments of insulin resistance, such as obtained by oral and intravenous glucose tolerance tests or the euglycemic hyperinsulinemic clamp54 55, and therefore elevated concentrations of proinsulin could be used as a marker of insulin resistance.

Etiological factors and pathophysiology It has been debated whether the etiology of insulin resistance and the meta-bolic syndrome is of genetic or environmental character.

It is known that obesity56, a sedentary life style57 and smoking58favor the development of insulin resistance. Dietary factors may induce oxidative stress59, a condition known to promote the development of insulin resistance 60 61, as well as improve insulin sensitivity by containing compounds with antioxidant activity62.

Low birth-weight, as a result of impaired growth in fetal life, has been as-sociated with components of the metabolic syndrome and type 2 diabetes mellitus63-65, and may be considered as an additional etiological explanation. Men with low birth-weight have been shown to have higher plasma cortisol concentrations compared to those with normal birth weight. Since glucocor-ticoid excess is known to induce hypertension, central obesity and glucose intolerance, it has been hypothesized that enhanced activity of cortisol could be the cause of insulin resistance and the metabolic syndrome66. It has also been shown that insulin resistance is associated with increased density of glucocorticoid receptors in skeletal muscle67.

13

Subjects with the metabolic syndrome have increased concentrations of inflammatory markers, such as CRP, fibrinogen and white blood cells, com-pared to those without the syndrome68. There appears to be a relation be-tween insulin resistance and CRP levels 69 70, and CRP has been shown to predict the development of type 2 diabetes71. Thus, it has been suggested that chronic inflammation may be a triggering factor for insulin resistance 72.

An increased activity of the sympathetic nervous system may also be of importance for the development of insulin resistance73. Low capillary density have been linked to insulin resistance74-76 implying that reduced peripheral blood flow may play a role in the development of insulin resistance.

However, it has also been shown that non-obese, non-diabetic relatives of subjects with type 2 diabetes are more insulin resistant than nondiabetic con-trols77, suggesting a genetic component. It is therefore most likely that the etiology of insulin resistance and the metabolic syndrome is a combination of both genetic and environmental factors.

The pathophysiological background of insulin resistance has been exten-sively examined. Defects in the actions of the insulin receptor, insulin recep-tor substrates, and glucose transport proteins78have all been reported as po-tential mechanisms causing a defect in insulin action. Endothelial dysfunc-tion could alter the presentation and function of insulin receptors and has been proposed to be a common factor in many of the features of the meta-bolic syndrome79.

Another pathophysiological component may be the fatty acid composition of the phospholipids of the skeletal muscle membranes that may affect the ability of the cell membrane to bind and transport insulin and glucose80. The fatty acid composition is related to insulin sensitivity81, and may be influ-enced by diet and physical activity. Genetic variations in enzyme actions involved in fatty acid metabolism may also be of etiological importance82.

In recent years also the triglyceride content in skeletal muscle83 and liver84 85 have been highlighted as drivers of the development of insulin re-sistance.

14

Coronary heart disease and the metabolic syndrome The presence of the metabolic syndrome has been associated with an in-creased risk of cardiovascular morbidity and mortality86 87and all the differ-ent metabolic disorders included in the syndrome have, by themselves, been associated with increased risk of CHD88-92.

Impaired glucose regulation Insulin resistance The association between insulin resistance and CHD has most often been examined by using fasting serum insulin as a proxy for insulin resistance. The role of insulin as a risk factor for CHD has been debated, and in a meta-analysis of 12 studies investigating this relationship it was concluded that hyperinsulinemia was only a weak risk indicator for the occurrence of car-diovascular disease93. Moreover, only a moderate correlation between plasma insulin and insulin sensitivity measured by the hyperinsulineamic euglycemic clamp method has been found94. However, it has been argued that the association between insulin levels and CHD could be confounded by the effects of other diseases resulting in malnutrition and weight loss and thus low insulin concentrations95.

There are few studies examining the association between CHD and insu-lin resistance measured by the hyperinsulineamic euglycemic clamp. One small study showed a significant association between the magnitude of insu-lin resistance and the degree of coronary artery disease96. A recent study from the ULSAM cohort showed that insulin resistance measured by the hyperinsulineamic euglycemic clamp, was a significant predictor of CHD independent of serum cholesterol, smoking and hypertension97.

The mechanisms behind the finding that insulin resistance increases the risk of CHD are not completely known. One explanation could be the strong association of hyperinsulinemia and insulin resistance with other metabolic risk factors included in the metabolic syndrome, such as dyslipidemia and increased PAI-1 activity. Insulin resistance could also be causative in itself, for instance by inducing impaired endothelial function98, smooth muscle proliferation99 and subsequent atherosclerosis. Moreover, increased levels of CRP are associated with the development of type 2 diabetes as well as CHD, which may suggest inflammation as a common basis for insulin resistance, the metabolic syndrome and atherosclerotic cardiovascular events68 100.

ProinsulinProinsulin is the precursor of insulin, and is processed in the beta cells of the pancreas to form insulin and C-peptide. In non-diabetic subjects proinsulin-like molecules (proinsulin and 31,32-split proinsulin) comprise about 10% of all insulin-like molecules101 in the peripheral blood, but in type 2 diabetes

15

mellitus the proportion is higher102. The effect of proinsulin on glucose me-tabolism is about 10%103 of that of insulin, but the half-life of proinsulin is longer (146 min versus 5 min) 104.

Proinsulin has recently been shown to be an independent predictor of CHD105 106 both in diabetic and non-diabetic populations, but the mecha-nisms for this association remain uncertain. Increased concentration of pro-insulin may be a marker of an underlying metabolic disturbance and thus not have a causative relationship with CHD105. However, clinical trials using proinsulin therapeutically for Type 2 diabetes resulted in several-fold higher incidence of MI in the treatment groups107. Elevated concentrations of proin-sulin and insulin are known to be strong determinants of PAI-1 activity108-110, and may thereby cause impaired fibrinolysis and increased risk of thrombosis. Moreover, in diabetic subjects elevated levels of PAI-1 have been found in extracted coronary atheroma111 and it has been proposed that PAI-1 might cause a reduction of vascular smooth muscle cells in the plaque and thus make it more vulnerable to rupture112. PAI-1 has also been associ-ated with components of the metabolic syndrome, such as insulin sensitivity and hypertriglyceridemia113. Thus, increased PAI-1 activity may be one potential link between proinsulin and coronary events.

Hyperglycemia Several studies have shown an association between the degree of hypergly-cemia and risk of micro-and macrovascular complications114 115. Increased blood glucose concentrations have been proven to impair endothelial func-tion in patients with type 2 diabetes mellitus, as well as in healthy subjects116117, and may thereby accelerate the atherosclerotic process.

However, in the United Kingdom Prospective Diabetes Study (UKPDS) intensive blood glucose control resulted only in a borderline significant re-duction of MI risk118. A possible explanation is that postprandial glycemia may be more strongly related to the risk of CHD than fasting plasma glu-cose, as suggested by the DECODE (Diabetes Epidemiology: Collaborative Analysis of Diagnostic Criteria in Europe) study group 119. An alternative explanation is the existence of proatherogenic metabolic disturbances al-ready in the prediabetic state120 121 that may contribute to the risk of macrovascular disease as much as clinical diabetes in itself.

DyslipidemiaAfter the results of the Scandinavian Simvastatin Survival Study (4S)122 in 1994 hypercholesterolemia, and especially elevated levels of LDL-cholesterol, are generally accepted as strong risk factors for CHD.

During the last years new parameters for measurement of lipid disorders have emerged. In a recent Swedish study apolipoptotein B (apoB), apolipo-protein A1 (apoA1) and the ratio between apoB/apoA1 were highly predic-

16

tive in the evaluation of cardiac risk. ApoB was a stronger predictor of risk than LDL-cholesterol. The importance of apolipoproteins was also valid in subjects older than 70 years and in subjects with concentrations of LDL-cholesterol below the median123. An explanation might be that apoB is pre-sent not only in LDL-particles, but also in the atherogenic VLDL124 and intermediate density lipoproteins. ApoA1, being present in HDL-particles, may play an important role in the atherosclerotic process especially in low-risk populations with normal cholesterol concentrations. ApoA1 has been shown to modify the ability of HDL-particles to remove cholesterol from cells125, and it may therefore be the content of apoA1 rather than cholesterol in HDL-particles that represents the atheroprotective capacity126. Moreover,it has recently been shown that treatment with a variant of apoA1 resulted in regression of coronary atherosclerosis compared to placebo127.

The most common lipid disorders in subjects with the metabolic syn-drome are raised serum triglyceride and lowered HDL-cholesterol concentra-tions while LDL-cholesterol concentrations may be normal. However, these individuals often have increased concentrations of small dense LDL-particles128, a condition that has been associated with an increased risk of CHD129, as well as an increased ratio between apoB and apoA1123 as signs of an atherogenic lipid profile.

The composition of fatty acids in cholesterol esters has been associated with cardiovascular disease130 131and with endothelial function. Palmitic (16:0) and palmitoleic (16:1) acids seem to impair endothelial function and might thereby be risk factors for CHD116 132.

HypertensionIt is well known that the incidence of CHD is increased in hypertensive pa-tients compared to normotensive controls133, also when hypertension is treated 89. Hypertension may promote the development of atherosclerosis by causing impaired endothelial function134-136, and may also be a trigger of acute plaque disruption by inducing mechanical stress on the arterial wall137138. Hypertension has also been associated with impaired fibrinolytic activ-ity, as evaluated by an impaired capacity for stimulated release of tissue plasminogen activator from vascular endothelium139, and may thereby pro-mote thrombosis formation.

Hypertension and insulin resistance During the last decades several studies have shown that a large part of pa-tients with hypertension are resistant to insulin-stimulated glucose uptake and are hyperinsulinemic compared to normotensive controls independently of obesity 140-142. It has also been shown that the prevalence of hypertriglyc-eridemia and low levels of HDL-cholesterol is increased in hypertensive subjects141.

17

The possibility has been raised that insulin resistance with compensatory hyperinsulinemia can cause and/or play a role in regulating hypertension. Insulin is known to promote sodium retention in the kidney143, to stimulate sympathetic nervous system activity144 145 and may be a stimulant of the growth of smooth muscle cells in the vessel wall99. However, not all patients with hypertension are insulin resistant 145. Insulin resistance has also been found in normotensive first-degree relatives of hypertensive patients, which might imply a genetic predisposition to insulin resistance in subjects prone to a high blood pressure.

Metabolic effects associated with antihypertensive treatment There is evidence that treatment with beta-blockers146and thiazide diuret-ics147 148 in hypertensive patients further reduce insulin sensitivity, thereby increasing the risk of developing type 2 diabetes mellitus149-152.

The magnitude of the reduction in insulin sensitivity varies between dif-ferent beta-blockers. Propranolol, atenolol and metroprolol have been shown to reduce insulin sensitivity by 25-32%, while with drugs with intrinsic ac-tivity, such as pindolol, the impairment was only 17%146 153.

The mechanism by which beta-blockers deteriorates insulin sensitivity is not fully understood, but several possibilities exist. One explanation is a reduced peripheral blood flow due to an increase in total peripheral vascular resistance154. It has been shown that vasodilating drugs, such as ACE-inhibitors155, alfa-blockers 156 and newer vasodilating beta-blockers157, can improve glucose metabolism. Beta-blockers may also lead to reduced insulin clearance resulting in hyperinsulinemia and subsequent down-regulation of insulin receptors and reduced insulin sensitivity. 158. Beta-blockers have also been associated with weight gain159, which may further reduce insulin sensi-tivity.

The mechanism by which thiazide diuretics reduce insulin sensitivity has been debated. One hypothesis is that potassium depletion causing impaired insulin secretion is the primary cause of impaired glucose tolerance seen during thiazide treatment160. An alternative explanation is that thiazides may increase circulating catecholamines and thereby decrease insulin sensitiv-ity161.

Diabetes mellitus and impaired glucose tolerance are both associated with an increased risk of CHD2 88, but it has been questioned whether these condi-tions when induced by beta-blockers and/or thiazides are associated with increased risk of CHD162.

During the last years beta-blockers and thiazide diuretics have been com-pared to newer antihypertensive drugs in large-scale studies. In the Captopril Prevention Project (CAPPP)163 there was no difference between treatments in preventing cardiovascular morbidity and mortality. However, significantly fewer patients developed diabetes in the captopril group compared to the group with beta-blockers/diuretics. In patients with diabetes at baseline there

18

was also a 66% lower incidence of MI in the captopril group compared to the group with beta-blockers/diuretics.

The Losartan Intervention For Endpoint reduction in hypertension study (LIFE)164 concluded that the angiotensin II blocker losartan prevented more cardiovascular morbidity and death than atenolol in hypertensives with left ventricular hypertrophy. There was also a lower rate of new-onset diabetes in the losartan group compared to the atenolol group.

The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart At-tack Trial (ALLHAT)165 showed that thizide diuretics were as good as ACE inhibitors and calcium channel blockers in preventing major forms of CHD even in subjects with diabetes. Also in this study there was a higher rate of new–onset diabetes in patients treated with diuretics compared with those treated with modern antihypertensive agents. However, the ethnic composi-tion of the ALLHAT study population differed from European populations and therefore the results may be difficult to apply in European countries.

The UKPDS166 also showed that antihypertensive treatment with atenolol was similarly effective in reducing the incidence of macrovascular complica-tions, such as MI, compared to captopril.

In conclusion, the results are diverging whether the metabolic abnormali-ties induced by beta-blockers and diuretics are of importance regarding fu-ture cardiovascular morbidity and mortality. However, the relatively short follow-up periods (3-5 years) that are common in large intervention studies may not be sufficient to detect effects on morbidity and mortality due to abnormal glucose metabolism induced by treatments with beta-blockers and thiazide diuretics.

19

Risk scores for coronary heart disease It is well known that the risk of developing CHD depends on a number of different risk factors2. Many of these risk factors interact and act synergisti-cally167 168, and it may therefore be difficult to assess the exact impact of each risk factor.

To grasp the overall picture of each patients´ risk of future CHD, several point scoring systems have been developed, and especially the Framingham score167 169 has been extensively used. Several studies have evaluated the performances of scores derived from the Framingham Heart Study in Euro-pean populations, and an overestimation of absolute risk of future CHD has been found 170-172. Recently new schemes based on the Framingham study47and the Prospective Cardiovascular Munster (PROCAM) Study172

have been presented, as well as a new European scoring system developed by the Score Project173 for use in clinical management of cardiovascular risk in European clinical practice. These new scoring schemes include traditional risk factors, such as blood pressure, serum cholesterol, smoking and diabe-tes.

However, during the last years new variables for prediction of CHD, such as the apolipoproteins and proinsulin have emerged. It has not been investi-gated if these markers of cardiovascular risk can improve the ability of a risk score in predicting future cardiovascular events.

20

AIMS

The specific aims of the studies were:

Paper ITo investigate the impact of variations in fasting blood glucose, blood pres-sure and BMI occurring between age 50 and 60 on the risk of developing MI during a follow-up of 17 years in a population-based sample of men. In the analysis, special attention was paid to subjects receiving antihypertensive therapy, with the hypothesis that drug-induced increase in fasting blood glu-cose would increase the risk of MI.

Paper II To examine differences in baseline risk factors between subjects developing stable angina pectoris demanding revascularisation (PTCA or CABG) with-out previous known MI, and subjects developing MI without previous known CHD during a 22 years follow-up period in a large cohort of middle-aged men.

Paper III To investigate the metabolic risk factors associated with the finding of a new Q/QS-pattern on the resting ECG in a population sample of 70-year-old men, and to investigate if the predictive power of the Q/QS-pattern was dependent of a history of MI or not during a follow-up of almost 10 years.

Paper IV To generate a new risk prediction score (the ULSAM score) including the apoB/apoA1 ratio and serum proinsulin, for the development of fatal and nonfatal MI in a population sample of middle-aged men, and to examine if these new risk factors for CHD could improve the predictive performance compared to schemes including only traditional risk factors, like the Fram-ingham and the PROCAM scoring schemes.

21

MATERIALS AND METHODS

SubjectsThe ULSAM study The Uppsala Longitudinal Study of Adult Men is a population-based study aimed at identifying risk factors for cardiovascular disease. Between 1970-73 all men born 1920-24 and resident in the municipality of Uppsala, Swe-den, were invited to participate in a health survey, and 2,322 out of the 2,841 invited men participated (82 %). All subjects gave informed consent. The local Ethics Committee of the Medical faculty at Uppsala University has approved the studies on several occasions.

Between 1980 and 1984 (at age 60) a re-examination of eligible subjects who had participated in the first survey was performed, in which 1,860 out of 2,130 subjects (87.5%) participated.

All eligible participants investigated in the first survey at age 50, traced by their ten-digit social security number, were invited to reinvestigation at age 70. The survey was carried out between August 1991 and May 1995. Participation rate was 73% (1,221 out of 1,681 invited) (figure1).

Eligible subjects have also been examined at age 77 and age 82, but those results are not included in the present thesis.

22

23222841AGE 50

AGE 60

AGE 70

2130 1860

1681 1221

Men invitedParticipants

98 dead, 94 moved

324 dead, 125 moved

460 nonparticipants

519 nonparticipants

270 nonparticipants

THE ULSAM COHORT

Figure 1. The ULSAM study

23

Study populations Paper I The study was based on the 1,860 subjects who participated both in the base-line investigations at age 50 and in the re-examination at age 60.

The population was grouped into subjects with (n=291) and without anti-hypertensive treatment (n=1,358) at age 60. Seventy-six subjects had mono-therapy with non-selective beta-blockers, 55 with selective beta-blockers, 63 with thiazide diuretics and 97 with a combination of beta-blockers and thi-azide diuretics. Forty-one subjects had hydralazine added to the treatment.

Subjects that had been hospitalised for MI before the examination at age 60 were excluded (n=75), as were subjects without MI but with angina pec-toris (according to in-hospital registers) during the follow-up period (n=115) and subjects lacking information about antihypertensive treatment at age 60 (n=21).

Paper II The study population was based on the 2,322 subjects who participated in the survey at age 50.

The population was divided into two groups; one group consisting of sub-jects who during the follow-up developed stable angina pectoris demanding revascularisation with PTCA or CABG without a preceding hospital diagno-sis of MI (n=70), and another group consisting of subjects who developed fatal or nonfatal hospital-treated MI during the follow-up, without previous hospital-documented manifestations of CHD (n=372). Also a control group without any known history of CHD during the follow-up period was identi-fied (n=1,701).

Subjects with a history of CHD (questionnaire and hospital files) before age 50 were excluded from the analysis (n=52), as well as subjects without MI, but with other manifestations of CHD (according to hospital files) dur-ing the follow-up, that did not require revascularisation (n=127).

24

Paper III The study was based on subjects that participated in baseline examinations and in the examination at age 70 (n=1,221). The population was divided into four groups according to ECG-status and MI/CHD diagnosis (according to in-hospital registers) at age 70.

The first group had Q/QS pattern (Minnesota code1.1/1.2/1.3) and MI di-agnosis (n=41). The second group had Q/QS but no MI or CHD diagnosis (n=91). The third group had MI diagnosis but no Q/QS (n=45), while the fourth group had normal ECG and no CHD diagnosis (controls, n=545).

Subjects with Q/QS pattern or left bundle branch block at age 50 (n=35), and subjects with pacemaker, left bundle branch block or no ECG registra-tion at age 70 were excluded (n=106). Subjects that did not fulfil the inclu-sion criteria of the four groups according to ECG-status and MI/CHD diag-nosis (ECG pathology other than Q/QS-pattern (Minnesota codes 4.1/4.2, 5.1-5.3 and 7.1) without MI diagnosis, normal ECG with CHD diagnosis other than MI, Q/QS-pattern with CHD diagnosis other than MI) were ex-cluded (n=358).

Paper IV The study was based on the 2,322 subjects who participated in the baseline investigations at age 50.

Subjects that had been hospitalised for, or had a history of CHD accord-ing to the questionnaire before these investigations, and subjects with base-line ECG showing ischemic signs (Minnesota code 1.1/1.2/1.3/4.1/4.2/5.1-5.3/7.1/9.2), were excluded (n=269). Subjects in the remaining population that lacked observations for any of the variables in the ULSAM score were also excluded (n=945). After this reduction in sample size the total data set consisted of 1,108 subjects. Fatal or non-fatal MI according to registry data was chosen as outcome variable (n=251 over 28.7 years of follow-up).

For cross-validation the total data set was divided into a training data set (subjects born 1921, 1922 and the second half of 1924, n=574, numbers of MI=135) in which the score was generated, and a prediction data set (sub-jects born 1920, 1923 and the first half of 1924, n=534, MI=116) for valida-tion of the score and for comparisons with the PROCAM and Framingham scores.

25

2322

1860

70 372 1701

462

52

2322

1649

291 1358

127

2322

1221

1101

722

41 91 45 545

358

2322

1108

269945

574 534

2143

PAPER I

PAPER II

PAPER III

PAPER IV

CHD before 50

Nonparticipants-60

Treatment No treatment

CHD, not revascularized

MI before 60Angina pectoris without MI

Nonparticipants-70

Other ECG/diagnosiscombinations

Non-evaluable ECG-70

CHD before 50Missing data

Training set Prediction set

Q+,MI+ Q+,MI- Q-,MI+ Controls

75115

21

Treatment NK

Revasc.angina MI Controls

106

35 Q/QS or LBBB at 50

Figure 2. Study populations in study I-IV. NK= not known. LBBB=left bundle branch block

26

Data Collection Investigations at age 50 AnthropometryHeight was measured to the nearest whole centimetres (cm) and weight (in undershorts) to the nearest whole kilogram (kg). BMI was calculated as weight (in kg) divided by height (in meters) squared.

Blood pressure Blood pressure was measured on the right arm after 10 minutes' rest in the recumbent position and after another 2 minutes in the sitting position. Mer-cury manometers (Kifa Ercameter, wall-model) were used. Systolic and dia-stolic blood pressures were read to the nearest 5 mmHg. The diastolic blood pressure was recorded at the disappearance of the Korotkoff sounds (phase five).

Glucose and Insulin Metabolism Blood glucose was measured by spectrophotometry using the glucose oxi-dase method. The intra-individual CV for fasting plasma glucose was 2.9 %.

An intravenous glucose tolerance test (IVGTT) was performed by injec-tion of a glucose dose of 0.5 g/kg of a 50% solution given into an antecubital vein during 2.5 minutes. Samples for the determination of blood glucose concentration were drawn before and 6, 20, 30, 40, 50 and 60 minutes after the start of the glucose injection. The serum insulin concentrations during the IVGTT were measured in duplicate of blood samples drawn before and 4, 6, 8 and 60 minutes after the start of the glucose injection. The serum in-sulin was determined with the Phadebas Insulin Test (Pharmacia AB, Upp-sala, Sweden).

Intact proinsulin and 32-33 split proinsulin concentrations were analysed using the two-site immunometric assay technique174 between 1995 and 1998 at the Department of Clinical Biochemistry, Addenbrooke's hospital, Cam-bridge, UK, using plasma samples that had been stored in liquid nitrogen for 15 years and thereafter stored frozen in -70°C.

Serum lipids Determinations of serum cholesterol and Tg concentrations were performed on a Technicon Auto Analyzer type II in 1981-82 on serum samples that had been stored in liquid nitrogen since 1970-73.

HDL-cholesterol was assayed in the supernatant after precipitation with a heparin/manganese-chloride solution. LDL-cholesterol was calculated using Friedewald's formula: LDL= serum cholesterol-HDL-(0.42xserum Tg).

27

The values presented are "Monarch adjusted", i.e. the values obtained were multiplied with a conversion factor for enabling comparison with the Monarch method used in the survey at age 70.

Apo(a) and apoB were determined by a two-site immuno-radiometric as-say and ApoA1 by a competitive radioimmunoassay in 1988, with use of commercial kits from Pharmacia (Uppsala, Sweden), in samples that had been stored in liquid nitrogen since sampling.

The percentage composition of methylated fatty acids (14:0) to (22:6) in serum cholesterol esters was determined by gas chromatography175.

Vitamin E In 1986 Alpha tocopherol (vitamin E) was determined by high performance liquid chromatography176.

ECGA standard 12-lead resting ECG was recorded at 50 mm/s and 10 mm/mV, including leads I, II, III, -aVR, aVL, aVF, and V1-6. The ECG machine was a direct-writing Mingograf 61 (Siemens-Elema Led, Solna, Sweden). The ECGs were classified according to the Minnesota Code177 by two experi-enced physicians at the Department of Clinical Physiology.

QuestionnaireA self-administered medical questionnaire was used to gather information on medical history. An interview was performed for completion of information regarding smoking habits, cause and age of death of parents and some psy-chosocial data.

Investigations at age 60 AnthropometryHeight was measured to the nearest whole cm and weight (in undershorts) to the nearest whole kg. BMI was defined as the weight (kg) divided by height (meter) squared.

Blood pressure The blood pressure was measured on the right arm in the recumbent position after 10 minutes rest by use of a mercury manometer (Kifa Ercameter, wall model) with a cuff width of 12.5 cm and a length of 35 cm. Korotkoff phase 5 was used for identification of diastolic blood pressure.

28

GlucoseBlood glucose was analysed by a glucose oxidase method (GOD-PERID, Boehringer Mannheim GFR), with use of an LKB 7400 photometer. A fast-ing blood glucose value was obtained in all participants.

DrugsInformation about medication was taken from the questionnaire and classi-fied according to the, at that time, current list of pharmaceutical specialties available in Sweden (FASS).

Investigations at age 70 AnthropometryHeight was measured to the nearest whole cm, and body weight to the near-est 0.1 kg. BMI was calculated as the ratio of the weight (in kg) to the height (in meters) squared.

Blood pressure Blood pressure was measured twice on the right arm to the nearest even fig-ure with the subject in the supine position after resting for 10 minutes. The mean of the two values are given. The cuff size was 12x35 cm or 15x45 cm depending on the arm circumference. Systolic and diastolic blood pressures were defined as Korotkoff phases I and V, respectively.

Oral glucose tolerance test (OGTT) An OGTT was performed where the subjects ingested 75 g glucose dissolved in 300 ml of water, and blood samples for plasma glucose and insulin were drawn immediately before, and 30, 60, 90, and 120 min after ingestion of glucose.

The area under the incremental curves for glucose and insulin were calcu-lated using the following formula: AUC=3(M30-M0+2(M60-M0)+2(M90-M0)+M120-M0) where M is the value of glucose or insulin at the specified time.

Early insulin response (insulinogenic index), as a measure of insulin se-cretion, was calculated as the ratio of the 30 min increment in insulin con-centration to the 30 min increment in glucose concentration.

Euglycaemic hyperinsulinaemic clamp The euglycaemic hyperinsulinaemic clamp technique was used to estimate in vivo sensitivity to insulin. The technique used was according to DeFronzo53,slightly modified.

Basal samples were taken 40 minutes after cannulation. Semisynthetic regular human insulin was infused in a primary dose for the first 10 minutes

29

and then as a continuous infusion (at 56 mU/min per body surface area) for 110 minutes to maintain steady state hyperinsulinaemia. The level of plasma glucose during the clamp study was maintained by measuring the plasma glucose every 5 minutes and adjusting the rate of infusion of a 20% glucose solution accordingly. The target plasma glucose level was 5.1 mmol/l. The glucose disposal (M) was calculated as the amount of glucose taken up dur-ing the last 60 minutes of the study and is given in mg/kg/min.

GlucosePlasma glucose in samples from the oral glucose tolerance test was measured by the glucose dehydrogenase method (Gluc-DH, Merck, Darmstadt, Ger-many). The intra-individual CV for fasting plasma glucose was 3.2%.

InsulinPlasma insulin from the oral glucose tolerance test and the clamp study was assayed using an enzymatic-immunological assay (Enzymmun, Boehringer Mannheim, Germany) performed in an ES300 automatic analyser (Boe-hringer Mannheim) and is given in mU/l.

Intact proinsulin and 32-33 split proinsulin concentrations were analysed using the two-site immunometric assay technique at the Department of Clinical Biochemistry, Addenbrooke's hospital, Cambridge, UK.

Lipids and apolipoproteins Cholesterol and Tg concentrations were analysed in serum and in the iso-lated lipoprotein fractions by enzymatic techniques using IL Test Cholesterol Trinders's Method and IL Test Enzymatic-colorimetric Method for use in a Monarch apparatus (Instrumentation Laboratories, Lexington, USA). HDL-particles were separated by precipitation with magnesium chlo-ride/phosphotumgstate. LDL-cholesterol was calculated using Friedewald's formula.

The apolipoproteins were analysed in a random subsample of 550 men. ApoB and apoA-l concentrations were determined by turbidimetry in a Monarch apparatus (Instrumentation Laboratories, Lexington, USA), using monospecifc polyclonal antibodies against apoB and A-l. Apo(a) was meas-ured by the Pharmacia Apo(a) RIA method (Pharmacia (a) RIA, Pharmacia Diagnostics AB, Uppsala, Sweden 1985).

30

ECG & Minnesota codes A standard 12-lead ECG was recorded at 50 mm/s and 10 mm/mV and evaluated according to the Minnesota code177by one experienced physician who was unaware of other data of the subjects. ECG status was classified as normal or abnormal on the basis of presence of one or more of the ECG di-agnoses listed below.

ECG diagnosis Minnesota code

Q or QS pattern 1.1

Q or QS pattern 1.2 or 1.3

High amplitude R waves 3.1 or 3.3

ST junction and segment depression 4.1 or 4.2

T-wave items 5.1 or 5.2 or 5.3

AV conduction defects 6.X

Left bundle branch block 7.1

Ventricular conduction defects 7.X

Atrial fibrillation or flutter 8.3

Arrhythmias 8.X

ST-segment elevation 9.2

Artificial pacemaker 6.8

Diseases & drugs Classification of drugs was performed according to the, at that time, current list of pharmaceutical specialties available in Sweden (FASS 1992/1993). All information about use of drugs was collected from the medical question-naire.

Hypertension prevalence was defined as a supine DBP of 95 mmHg or greater and/or treatment with anti-hypertensive drugs for hypertension. Men treated with these drugs for other reasons, i.e. cardiac failure, are thus not included in this definition.

Hyperlipidaemia prevalence was classified as serum cholesterol >6.5 mmol/l and/or serum triglycerides >2.3 mmol/l and/or lipid-lowering medi-cations.

Diabetes was diagnosed if the 120 min and one or more of the 30-90 min plasma glucose values at the OGTT were greater or equal to 11.1. mmol/l178

and/or antidiabetic treatment .

QuestionnaireTwo self-administered optically readable questionnaires were used. One was a questionnaire on general and medical background and the other concerned living conditions.

31

Registry data The cause of death registryMortality data was achieved from the cause of death registry. The register is updated yearly and contains personal identification number, home town, underlying cause of death, nature of the injury, multiple causes of death, date of death, basis for statement of cause of death, sex and age. The international classification code for diseases (ICD-9 and 10) is used for classification of the causes of death.

The hospital discharge registry Information on hospitalisations was achieved from the hospital discharge registry (HDR). The medical data in the register includes main diagnosis, secondary diagnoses, external causes of injury and poisoning and surgical procedures. The register is updated annually.

Definitions CHD was defined as ICD-9 codes 410-414, or ICD-10 codes I 20-25, MI as ICD 410/ I 21, angina pectoris as ICD 413/I 20 and cardiovascular diseases as ICD 390-459/I 00-99.

Data concerning subjects who underwent PTCA and CABG during the follow-up according to hospital files was received from the Departments of Cardiology and Thoracic surgery.

During 1970 to 2001 there were 499 registered fatal and nonfatal MI in the cohort, out of which 115 occurred during the first decade (incidence 5%), 195 the second decade (incidence 9%) and 189 during the third decade (inci-dence 10%).

32

Statistical analyses General analysis The statistical analyses were made using JMP and STATA (paper IV) statis-tical softwares. A p-value <0.05 was considered as statistically significant.

Non-normally distributed variables tested by Shapiro-Wilk´s test (w<0.95) were log-transformed before being analysed.

Differences between groups regarding metabolic characteristics were evaluated by factorial one-way ANOVA, or the chi squared test for differ-ences in proportions.

The independent power of the variables was evaluated by stepwise multi-ple Cox proportional hazard analysis (backward selection). Z-transformed standardized variables were used in the Cox proportional hazard analysis in order to make hazard ratios comparable. Hazard ratios are given for a one standard deviation change in the continuous variables with their 95% confi-dence intervals.

Specific analyses Paper I The risk associated with a 10% increase in blood glucose was calculated as follows: Hazard ratio**(log(1.1)/standard deviation).

Paper II Bootstrapping Since the group with revascularised angina pectoris and the group with MI without previous known CHD were not independent of each other a boot-strap method was used to test the significance of the differences in hazard ratios between the groups.

Bootstrapping is a method in which a new reference population is created from the collected data by performing a large number of random samples from the original data. A random sample including an equal amount of ob-servations as the original sample is identified and adequate calculations are performed. The sample is then returned to total dataset and the procedure is repeated 1000-10000 times. In this new reference population the standard errors were estimated and the p-values for the differences between hazard ratios were calculated.

The expression “bootstrap” is derived from the tale of baron von Münch-hausen who managed to get himself up from a marsh by lifting himself in his bootstraps.

33

Paper III Differences between groups (Q/QS-pattern versus no Q/QS-pattern, and MI diagnosis versus no MI diagnosis) regarding metabolic characteristics were evaluated by factorial two-way ANOVA. A significant interaction term be-tween these two factors was regarded as an evidence for an impact of MI diagnosis on the risk factor profile in subjects with a Q/QS-pattern.

Paper IV Generation of risk prediction score For cross-validation the total data set was divided into a training data set in which the score was generated, and a prediction data set for validation of the score and for comparisons with the PROCAM and Framingham scores.

The multiple Cox Proportional Hazard Analysis was used for identifying independent baseline risk factors for fatal or nonfatal MI during the follow-up in the training data set. The continuous variables (serum proinsulin, sys-tolic blood pressure and apoB/apoA1 ratio) were then divided into decen-tiles. The risk of MI during the follow-up was calculated for each decentile, and cut-off levels for increasing risk were decided. To calculate the scores between the variables and within each variable category, the decentile-groups were dichotomised and entered into a logistic regression model to-gether with family history of MI and smoking. The parameter estimates cal-culated in this manner were rounded to whole numbers and used as scoring points. The maximal score that could be achieved was 26 points.

Validation of the score The estimated risk of fatal and nonfatal MI calculated by use of the ULSAM score was compared to the observed event rate over the entire range of de-centiles of risk, using the Hosmer-Lemeshow goodness of fit statistics179.

Areas under receiver operating curves180 181 were used to measure the abilities of the ULSAM, Framingham and PROCAM scores to separate those who developed fatal and nonfatal MI from those who did not in the predic-tion data set.

34

Discussion of Methods Selection bias; general aspects ParticipationThe participation rates were 82 %, 87.5% and 73% at the examinations at age 50, 60 and 70. It is known that non-participants in cohort studies gener-ally have more social and alcoholic problems and a higher morbidity in CHD compared with participants182. Thus, estimated risks of CHD may be under- rather than overestimated. However, higher participation rates than those in ULSAM are rarely found in population studies.

Mortality and morbidity analyses No subjects were lost to follow-up due to missing data thanks to the official hospital and cause-of-death registries in Sweden. The cause of death registry includes all deaths of persons registered in Sweden at the time of death; deaths occurring outside Sweden are thus also recorded.

All hospital admissions in the Uppsala health care region have been re-ported to the HDR since 1965. Since 1984 the reporting to the HDR has been mandatory in Sweden and from 1987, the HDR covers all public, in-patient care in Sweden.

The validity of the HDR concerning MI diagnosis has been investigated by the Center of Epidemiology at the National Board of Health and Welfare. It was found that the MI diagnosis was correct in 95% and missed in 3% of the cases183. The discharge register does not include non-fatal home-treated or unrecognised MI. Unrecognised MI have been estimated to constitute 18-35% of the total number of cases of MI11-16 and was not included as cases in the present study, as in most other cohort studies of MI. To study unrecog-nised MI was however the scope of paper III.

ProinsulinDue to a freezer failure proinsulin at age 50 was analysed in a random sam-ple of 1,335 out of the total 2,322 subjects at age 50. Where appropriate, the statistical analyses were also performed only in subjects with proinsulin val-ues. This reduction in sample size did not substantially change the results (paper I, II). In paper IV, in which proinsulin was a focus, the size of the cohort was however substantially reduced.

The proinsulin analyses at age 50 were performed in samples that had been stored in liquid nitrogen for 15 years and thereafter stored frozen in -70°C until the analyses in 1995-98. It cannot be ruled out that long time storage may have influenced the absolute values of proinsulin. However, it would have affected all samples in the same manner independently of the subjects developing CHD or not.

35

Smoking Ex-smokers were classified as non-smokers since there were no differences in risk of fatal and nonfatal MI (p=0.52) or total mortality (p=0.56) between non-smokers and ex-smokers in the ULSAM cohort.

Selection bias; specific aspects Paper IIThe aim of the study was to evaluate differences between subjects develop-ing stable angina pectoris without previous known MI and subjects develop-ing MI without previous known CHD. We chose to study angina patients demanding CABG or PTCA in order to achieve an objective criterion of angina pectoris, as many false positive cases otherwise would be identified by medical history only. By choosing this approach we may have selected an angina pectoris group with a more severe form of arteriosclerosis than the average angina pectoris patient.

On the other hand, subjects with metabolic disorders, such as diabetes, may have been denied surgery because of a high intraoperative risk, and thus the subjects in the revascularised group may have had less metabolic abnor-malities compared to the average angina patient and compared to the sub-jects in the group with MI. However, the prevalence of diabetes (fasting blood glucose >6.1 mmol/l) at the examination closest to the events (exami-nation at age 60) did not differ between the revascularised group and the group with MI (7% versus 11%, p=0.30).

Although the angina pectoris patients did not have hospital diagnosis of MI prior to the revascularisation it cannot be excluded that they had experi-enced an unrecognised MI. Likewise, it cannot be excluded that the MI cases could have had a history of angina pectoris not rendering a hospital diagno-sis. Thus, several types of selection bias might occur in the study population in paper II. It is however very hard to perform a similar population based study with a long follow-up period without selection bias and without risk of diluting an angina group with MI cases and vice versa.

Since 23 subjects in the revascularised group with angina pectoris had un-stable angina pectoris according to hospital files, the analyses were also per-formed without these subjects. However, this exclusion did not substantially change the results.

36

Paper IV Proinsulin and apoB/apoA1 were analysed only in 1,108 out of 2,053 eligi-ble subjects, which may have induced a selection bias.

The incidence of MI was significantly higher in those with observations for proinsulin and the apoB/apoA1 ratio compared to those without observa-tions. However, there were no differences concerning traditional atherogenic risk factors between subjects with MI in the analysed group and subjects with MI in the excluded group, except for higher total cholesterol (7.3 versus 6.9 mmol/l, p=0.002) in those with observations. Thus, the subjects with MI in the analysed population were representative of the total MI-population.

The training and the prediction data sets did not differ in traditional car-diovascular risk factors.

The predictive power of CHD risk factors during long follow-ups The assessment of the ability of different risk factors in predicting CHD morbidity and mortality over a long follow-up was based on single meas-urements at baseline. It is known that the risk factor status in individuals at baseline may change during the course of follow-up184, and so the length of follow-up may influence the predictive power.

Previous studies have concluded that serum cholesterol is a strong long-term risk factor for CHD, while the long term effect of smoking is likely to be dependant on the rate of giving up smoking during the follow-up. Blood pressure has also been shown to be a strong predictor of CHD even if it tends to loose its predictive power after 15-20 years185-187. HDL-cholesterol, physical activity, diabetes and parental history all remained predictive of CHD over 10-15 years in a British study188, while variables with less reli-ability, such as serum insulin, were not predictive over a longer follow-up.

37

RESULTS AND DISCUSSION

Paper I ResultsIn the group with antihypertensive medication at age 60 the LDL-to HDL-cholesterol ratio at age 50 and an increase in fasting blood glucose between age 50 and 60 were significant independent risk factors for developing MI after the age of 60 over the 17.4 year follow-up. An increase of 10% in fast-ing blood glucose in the treatment group was associated with a 21.7% in-creased risk of MI after age 60. After addition of proinsulin to the model, the impact of increase in blood glucose declined, but was still a significant pre-dictor of future MI (Figure 3a).

1.0 1.2 1.4 1.6HAZARD RATIO

1.8

Change in fasting blood glucosebetween age 50 and 60

LDL/HDL-cholesterol ratio

Proinsulin

= Without proinsulin

= With proinsulin

Figure 3a. Hazard ratios for MI after age 60 in the group with antihypertensive treatment, with and without proinsulin in the model. Stepwise multivariate CoxPro-portional Hazard analysis.

38

In the group without antihypertensive medication the LDL- to HDL-cholesterol ratio and serum proinsulin at age 50, as well as increase in sys-tolic blood pressure between age 50 and 60 were independent predictors of subsequent MI, but increase in fasting blood glucose was not a risk factor in this group (Figure 3b).


1.8

Change in blood glucose

LDL/HDL cholesterol ratio

Proinsulin

= Without proinsulin

= With proinsulin

0.8

Change in systolic blood pressure

Figure 3b. Hazard ratios for MI after age 60 in the group without antihypertensive treatment, with and without proinsulin in the model. Stepwise multivariate Cox Proportional Hazard analysis.

39

There was a significant interaction between proinsulin and antihyperten-sive treatment regarding increase in fasting blood glucose. Thus, those with the highest proinsulin concentrations at baseline and antihypertensive treat-ment at age 60 showed the greatest increases in fasting blood glucose con-centrations between the age of 50 and 60 (Figure 4).

-0,4-0,2

00,20,40,60,8

11,21,41,6

Serum proinsulin quartiles

Cha

nge

in b

lood

glu

cose

(mm

ol/l)

treateduntreated

2 3 41

Figure 4. Mean and standard error of the mean for change in blood glucose between age 50 and 60 according to serum proinsulin levels at age 50 in the group with, and the group without antihypertensive treatment at age 60.

There was a significant relation between the change in fasting blood glu-cose and the average of fasting blood glucose at age 50 and 60, which indi-cated that there was no regression towards the mean.

40

DiscussionDuring the last decade several studies have shown that a large part of pa-tients with hypertension are resistant to insulin-stimulated glucose uptake and are hyperinsulinemic compared to normotensive controls140-142. In pre-vious prospective cohort studies treatment with beta-blockers and/or diuret-ics was associated with increased risk for diabetes149-151. This is probably due to increased insulin resistance during treatment with beta-blockers and thiazides146-148.

The relation of insulin resistance to CHD has been well-established189.However, the influence of the metabolic changes induced by antihyperten-sive treatment on the risk of MI, has been questioned162.

In the present study increase in fasting blood glucose, as well as the LDL- to HDL-cholesterol ratio and serum proinsulin at baseline, were significant risk factors for MI after age 60 in the group with antihypertensive treatment. In spite of the treatment, the blood pressure at age 60 was not normalised, but neither baseline nor achieved blood pressure at age 60 were related to the incidence of MI.

In the multivariate analysis, the impact of increase in fasting blood glu-cose on the risk of MI was independent of baseline lipids, fasting insulin and glucose, BMI and blood pressure. However, when proinsulin was added to the models, the predictive power of increase in fasting blood glucose de-clined, indicating that insulin resistance at baseline, for which proinsulin may be considered as a marker54 55, explains a part of the predictive power of the induced increase in fasting blood glucose in hypertensives.

However, even when the impact of insulin resistance was taken into ac-count, the deleterious effect of the increase in fasting glucose was still sig-nificant, suggesting that both insulin resistance at baseline and the induced increase in insulin resistance caused by antihypertensive treatment are risk factors for MI.

An additional explanation of the deleterious effect of the increase blood glucose is the finding that glucose may directly affect the development of arteriosclerosis by impairing endothelial function. It is known that patients with type 2 diabetes mellitus have impaired endothelial function190 191, and increased blood glucose concentrations have been proven to impair endothe-lial function in healthy subjects as well116 117.

41

Paper II ResultsSerum lipids, as well as the apolipoproteins, were the most powerful risk factors for later revascularisation because of angina pectoris without previ-ous hospital diagnosis of MI. Neither blood pressure, nor BMI, insulin or glucose parameters at IVGTT were significant risk factors in this group.

Serum lipids and apolipoproteins were also important predictors of MI without previous hospital diagnosis of CHD. However, in this group also smoking, blood pressure, BMI, insulin and glucose parameters at IVGTT, serum proinsulin, as well as palmitic, palmitoleic, stearic and oleic acid were powerful risk predictors.

In stepwise multivariate Cox proportional hazard models baseline values of LDL- and HDL-cholesterol (protective) turned out to be independent risk factors in subjects with angina pectoris. In the group with MI, the same analysis showed that serum proinsulin, SBP, serum apolipoprotein B, smok-ing and HDL-cholesterol (protective) were independent predictors of MI during the follow-up period. When comparing hazard ratios between the two groups, there were significant differences in hazard ratios for DBP, serum proinsulin and split proinsulin, but not for HDL- or LDL-cholesterol (Figure 5).

00,20,40,60,8

11,21,41,61,8

22,2

LDL HDL Proinsulin DBP

Haz

ard

ratio

Angina

MI

p< 0.05

p< 0.05

Figure 5. Differences in hazard ratios for a one standard deviation variation in car-diovascular risk factors and their 95 % confidence intervals, between subjects revas-cularised because of angina pectoris without prior MI and subjects with MI without prior CHD.

The hazard ratios for metabolic characteristics at age 50 were also calcu-lated without subjects with diabetes at age 60 (the examination closest to the events), but the results did not change.

42

DiscussionThe main cause of acute coronary syndromes is plaque disruption with su-perimposed thrombosis33-35, and it may be the vulnerability of the plaque rather than the extent of coronary atherosclerosis that is the major determi-nant of acute MI38. This vulnerability has been extensively examined in sev-eral studies and inflammatory cell activity, cigarette smoking, disturbed fi-brinolysis and hemodynamic factors38 40 41 112 138have been suggested to be potential risk factors for plaque rupture and thrombosis.

It is therefore not surprising that we found both similarities (lipid profile) and differences (serum proinsulin and DBP) between subjects developing acute MI without prior known CHD and subjects with angina pectoris de-manding revascularisation without prior known MI, regarding the risk factor profile at baseline.

Insulin resistance113 192, as well as elevated concentrations of proinsulin and insulin 108-110, are known to be strong determinants of PAI-1 activity, and may thereby cause impaired fibrinolysis and increased risk of thrombo-sis. In diabetic subjects elevated levels of PAI-1 has been found in extracted coronary atheroma111and it has been proposed that PAI-1 might cause a re-duction of vascular smooth muscle cells in the plaque and thus make it more vulnerable to rupture. Thus, our finding that proinsulin is a risk factor for MI, but not for angina pectoris may be partly explained by increased PAI-1 activity.

In the present study elevated blood pressure was an independent risk fac-tor for MI. It is well known that the incidence of CHD is increased in hyper-tensive patients compared to normotensive controls133, also when the hyper-tension is under treatment 89. Hypertension may promote the development of atherosclerosis by causing impaired endothelial function134-136, and may also be a trigger of acute plaque disruption by inducing mechanical stress on the arterial wall137 138. Hypertension has also been associated with impaired fibrinolytic activity, as evaluated by an impaired capacity for stimulated release of tissue plasminogen activator from vascular endothelium139, and may thereby promote thrombosis formation.

43

PAPER III ResultsImpact of Q/QS-pattern in cross-sectional analysis Subjects with Q/QS-pattern at age 70 had lower values for early insulin re-sponse and total plasma insulin AUC at the 2-hour OGTT compared to those without Q/QS. This difference seemed to be caused by lower values in the group with Q/QS without MI diagnosis, but the interaction term between Q/QS-pattern and MI diagnosis was not significant. Insulin sensitivity (M-value at the euglycaemic hyperinsulinaemic clamp) and other risk factors analysed did not differ significantly between those with and without Q/QS-pattern.

Impact of history of MI in cross-sectional analysis The subjects with MI diagnosis at age 70 had significantly higher concentra-tions of serum proinsulin, triglycerides, apolipoproteins and HDL-cholesterol, but lower blood pressure when compared to those without MI diagnosis. There were no differences in total and LDL-cholesterol between the groups.

44

Prevalence of different risk factors at age 70 The population was divided into 4 groups according to ECG-status and MI/CHD diagnosis at age 70.

While the prevalence of diabetes was elevated to a similar degree in all the groups with Q/QS pattern and/or MI diagnosis compared to the control group with normal ECG and no CHD diagnosis, an increased prevalence of smoking was only seen in those with Q/QS-pattern without MI diagnosis. The prevalence of hyperlipidemia was increased in the two groups with MI diagnosis, but not in the group with Q/QS-pattern without MI diagnosis. There was no significant difference in hypertension prevalence between the groups (Figure 6).

0

10

20

30

40

50

60

70

Hypert

ensio

n

Hyperl

ipide

mia

Diabete

s

Smoking

%

Q+/MI+Q+/MI-Q-/MI+Controls

Figure 6. Prevalence of risk factors at age 70.

45

Follow-up after age 70 Q/QS-pattern was a significant predictor of total and cardiovascular mortal-ity both in the univariate and the multivariate Cox proportional hazard mod-els when the analysis were adjusted for smoking, hypertension, diabetes mellitus, LDL-cholesterol and MI diagnosis (Figure 7a and b).

MI diagnosis was a significant predictor of cardiovascular and total mor-tality in the univariate analysis, but lost its significance when the variableQ/QS-pattern was included in the model.


1.8

MI diagnosis

Hypertension

Smoking

Q/QS-pattern

Diabetes

0.80

LDL cholesterol

Figure 7a. Hazard ratios for total mortality after age 70. Stepwise multivariate Cox Proportional Hazard analysis.


1.8

MI diagnosis

Hypertension

Smoking

Q/QS-pattern

Diabetes

0.80

LDL cholesterol

2.0

Figure 7b. Hazard ratios for cardiovascular mortality after age 70. Stepwise multi-variate Cox Proportional Hazard analysis.

46

There was a significant difference concerning total and cardiovascular mortality after age 70 between the four groups. In the group with Q/QS without MI diagnosis the incidence of total and cardiovascular mortality was significantly higher compared to the control group in a post-hoc analysis. The highest mortality rates were found in the group with Q/QS and MI diag-nosis, but the interaction term between the two variables were not significant in the Cox proportional hazard analysis (Figure 8).

Q/QS No Q/QS

No MI diagnosisMI diagnosis

05

1015202530

% Total mortality

Q/QS No Q/QS

No MI diagnosisMI diagnosis

05

1015202530

% CVD mortality

21/91

12/41

8/45

68/545

11/91

11/41

26/545

2/45

Figure 8. Total and cardiovascular mortality in the population during 9.4 years of follow-up after age 70 divided by to the presence or absence of Q/QS pattern and MI diagnosis at age 70.

47

DiscussionIn accordance with previous studies of the prognostic value of ECG abnor-malities23-32, the present study showed that the finding of a new Q/QS-pattern on the resting ECG was a significant predictor of cardiovascular and total mortality. The prognostic importance remained significant when the models were adjusted for smoking, diabetes, hypertension, hyperlipidemia and MI diagnosis. When looking at the subgroup with Q/QS-pattern without MI diagnosis, the risk of total and cardiovascular mortality was significantly increased compared to a control group without known CHD, further under-lining that an unexpected finding of a new Q/QS on the ECG is not an inno-cent sign11-16.

The major metabolic disturbances in subjects with Q/QS pattern without MI diagnosis were a reduced early insulin response and reduced plasma in-sulin AUC at the OGTT, as may be seen in the progressive development of type 2 diabetes mellitus193 194. These findings are in accordance with previ-ous studies that have indicated an association between diabetes mellitus and unrecognised MI15 17 18 and it has been proposed that autonomic neuropathy due to diabetes may explain this association19 20. However, others have failed to prove that diabetics are more prone to unrecognised MI than non-diabetics21 22.

Even though the Q/QS pattern was associated with indices of diabetes de-velopment in the present study, it was a significant predictor of total and cardiovascular mortality even when the effect of diabetes was accounted for in the multiple analyses. Thus, it is not only the diabetic state per se that causes the increased mortality in these subjects. It seems rather that the Q/QS pattern indicates myocardial damage that may lead to heart failure or conductance abnormalities, and thus increased risk of premature death.

Another contributory cause to the increased risk of total and cardiovascu-lar mortality in subjects with Q/QS-pattern may be the high percentage of smokers seen in the group with Q/QS pattern without MI diagnosis com-pared to the control group. It is well known that smoking accelerates the atherosclerotic process and thus leads to an increased risk of coronary events195. Previous studies have shown that patients with unrecognised MI tend to avoid contacts with physicians11and hence reduce the possibility of getting anti-smoking advices. However, the predictive power of Q/QS pat-tern was independent of smoking in the multiple analyses.

The high percentage of smokers could also be an explanation to the low-ered insulin response seen in this group. Smokers tend to have a lower BMI196 197, and thereby lower insulin concentrations compared to non-smokers. However, when the analyses were adjusted for smoking, subjects with Q/QS-pattern still had significantly lower values for the insulinogenic index and total plasma insulin AUC at the 2-hour OGTT compared to those without Q/QS.

48

On the contrary, subjects with MI diagnosis did not have an increased percentage of smokers compared to controls, and had significantly lower blood pressure and no difference in LDL-cholesterol compared to those without MI diagnosis. It is likely that subjects with MI diagnosis participated in some form of rehabilitation program after the acute event, and thus may have modified their risk factor patterns. This may also have influenced the fact that MI was not an independent predictor of cardiovascular and total mortality in the multiple analyses.

Paper IV ResultsGeneration of a risk prediction score (ULSAM score) The score was based on 251 cases of fatal and nonfatal MI within 28 years of follow-up in 1,108 men who were free of CHD at baseline. Serum proinsu-lin, the apoB/apoA1 ratio, SBP, smoking and family history of MI were found to be independent significant predictors of fatal and nonfatal MI in the training data set and were thus included in the score.

Validation of the ULSAM score There were highly significant differences regarding scoring points of the ULSAM score between subjects who developed MI during the follow-up and subjects who did not in both the training and the prediction datasets.

The estimated risk of fatal and nonfatal MI calculated by use of the UL-SAM score was compared to the observed event rate over the entire range of decentiles of risk, using the Hosmer-Lemeshow goodness of fit statistics179.The estimated risk showed a good agreement with the observed incidence.

Figure 9 presents receiver-operating curves180 181 measuring the abilities of the ULSAM, Framingham and PROCAM scores to separate those who developed fatal and nonfatal MI from those who did not in the prediction data set during the follow-up of 28 years. Although the ULSAM score showed the best performance the areas under the curves did not differ sig-nificantly between the scores198.

49

1 - Specificity1.00

Sens

itivi

ty

1 - Specificity0.00 0.25 0.50 0.75

0.25

0.50

0.75

1.00

ULSAM 66%

FRAMINGHAM 61%

PROCAM 63%

Figure 9. Receiver operating curves showing performance of ULSAM, PROCAM and Framingham scores in predicting risk of MI within 28 years of follow-up among middle-aged men in the ULSAM cohort. Numbers refers to areas under the curves.

In order to demonstrate the power of the study the 95% confidence inter-vals for the pairwise differences of areas under receiver-operating curves (expressed as percentages of the mean area) were calculated. These intervals did not exceed 12% of the mean area. These intervals were interpreted as being narrow indicating that the power to detect large differences was high.

DiscussionSerum proinsulin, the ratio apoB/apoA1, smoking, SBP and family history of MI were independent predictors of fatal and non-fatal MI over a period of 28 years in the ULSAM cohort, and were therefore included in the scoring scheme. Traditional risk predictors such as LDL- and HDL- cholesterol, serum triglycerides and diabetes prevalence, were also included in the multi-ple Cox proportional hazard analysis, but these variables were not independ-ent predictors of MI when apoB/apoA1 and serum proinsulin were added to the analysis.

Hypercholesterolemia and especially high concentrations of LDL-cholesterol are generally accepted as strong risk factors of cardiovascular disease122. However, several studies have suggested that apoB and apoA1 also are important risk predictors126 199-201. A recent large Swedish study showed that apoB, apoA1 and especially the ratio of apoB/apoA1 were inde-

50

pendent risk predictors of fatal MI after adjustment for age, total cholesterol and triglycerides. The importance of apolipoproteins was also valid for sub-jects older than 70 years and for subjects with concentrations of LDL-cholesterol below the median123. ApoB is present not only in LDL-particles, but also in VLDL124 and intermediate density lipoproteins, and may there-fore be a better indicator of the total atherogenic content of the lipoproteins. ApoA1 has been shown to modify the ability of HDL-particles to remove cholesterol from cells125 and may play an important role in the atheroscle-rotic process especially in low-risk populations with normal cholesterol con-centrations, and it may therefore be the content of apoA1 rather than choles-terol in HDL-particles that is of most importance for the development of atherosclerosis126.

The role of insulin as a risk factor for CHD has been debated, and in a meta-analysis of 12 studies investigating this relationship, it was concluded that hyperinsulinemia was only a weak risk indicator for the occurrence of cardiovascular disease93. Proinsulin, on the other hand has recently been shown to be an independent predictor of CHD105 106, even though the mecha-nisms for this association remain uncertain. Insulin resistance113 192 and ele-vated concentrations of proinsulin and insulin 108-110 are known to be strong determinants of PAI-1 activity, and may thereby cause impaired fibrinolysis and increased risk of thrombosis. Thus, increased PAI-1 activity may be one potential link between proinsulin and coronary events.

In the present study, the performance of the ULSAM score, including pro-insulin and the apoB/apoA1 ratio, in predicting the absolute risk of fatal and nonfatal MI was good. The ability of the score to separate those who devel-oped fatal and non fatal MI from those who did not during the follow-up (measured as areas under receiver operating curves) was at least as good as the Framingham and PROCAM scores. Thus, newer risk predictors, such as proinsulin and the apoB/apoA1 ratio, seem to contain at least as much pre-dictive power as established risk factors, such as different glucose parame-ters, HDL- and LDL-cholesterol and serum triglycerides, and might there-fore be used more frequently in the future.

51

GENERAL DISCUSSION

Metabolic consequences of antihypertensive treatment Antihypertensive treatment reduces the risk of CHD by 14% according to a meta-analysis performed in 1990202. However, in the beginning of the era of antihypertensive treatment an even larger reduction in CHD morbidity and mortality was expected.

One of the reasons for this rather modest effect may be the metabolic dis-turbances induced by older antihypertensive drugs, such as beta-blockers and thiazide diuretics. During the 1980s it was shown in several studies that in-dividuals with hypertension were insulin resistant compared to normotensive controls140-142 and that treatment with beta-blockers and thiazide diuretics induced decreased insulin sensitivity146 147 and thus increased risk of devel-oping type 2 diabetes mellitus149-152. This has also been confirmed in a re-cent study in which treatment with thiazide diuretics and beta-blockers was associated with an aggravated metabolic profile including an increase in the ratio between apolipoprotein B and apolipoprotein A1148.

The clinical importance of the induced metabolic abnormalities has been questioned. In a Swedish study162 it was found that development of diabetes mellitus related to antihypertensive treatment did not increase the risk of coronary events. However, that study used diabetes mellitus (fasting blood glucose>7.0 mmol/l) as a categorical variable, and since the relationship between blood glucose concentrations and risk of CHD appears to be lin-ear203, a categorical grouping based on fasting blood glucose diminishes the statistical capacity of detecting an effect. Furthermore, the confidence inter-val of the risk associated with drug related diabetes mellitus was wide. It is therefore hazardous to exclude an effect of drug related diabetes mellitus on the risk of MI based on the data from that study.

In the ULSAM cohort increase in fasting blood glucose during 10 years was an independent risk factor for MI in 291 men with antihypertensive treatment with beta-blockers and thiazide diuretics during a follow-up of 17 years (paper I). There was an interaction between insulin resistance (indi-cated by the serum proinsulin concentration) at baseline and antihypertensive treatment on the degree of increase in blood glucose. However, increase in fasting blood glucose was an independent risk factor also when the analysis was adjusted for insulin resistance at baseline.

52

A limitation of the study was that the comparisons were made between one group with metabolic abnormalities at baseline at age 50 (the group with treatment) and one group without such abnormalities (the untreated group). Subjects without antihypertensive treatment at age 60 but with insulin resis-tance (the highest quartile of serum proinsulin) at baseline also had an in-crease in blood glucose during 10 years as a sign of an increase in insulin resistance (Figure 4). However, this increase was significantly smaller than in the group with antihypertensive treatment, indicating that insulin resis-tance at baseline only explained parts of the increase in blood glucose in the treatment group.

To further examine the influence of pre-existing risk factors on increase in blood glucose, a subgroup with hypertension ( 154/93 mmHg) but with-out antihypertensive treatment at age 60 was identified (n=166). This group did not differ from the group with antihypertensive treatment concerning metabolic abnormalities at baseline, but in contrast to the treated group these subjects did not increase in blood glucose during the following 10 years.

During the 1990s several large intervention studies were performed where older antihypertensive drugs were compared to newer agents such as calcium channel blockers, ACE-inhibitors and angiotensin II (AII) blockers. In sev-eral of these studies (CAPPP163; LIFE164; ALLHAT165) there was an in-creased incidence of new-onset diabetes in subjects treated with beta-blockers/diuretics compared with those treated with more modern drugs. In the HOPE trial204 ACE inhibition was even associated with less new-onset diabetes in a population at high cardiovascular risk when compared to pla-cebo. In the LIFE and ANBP2205 studies treatments with ACE-inhibitors/AII-blockers were associated with a lower risk of cardiovascular events compared to older agents, while in the CAPPP, ALLHAT and UKPDS studies older drugs performed at least as well as newer agents in preventing cardiovascular endpoints.

A major limitation of these studies is that the follow-ups are limited to about 5 years. A prolongation of two large intervention studies206 207 showed that a follow-up of at least 8-10 years is necessary to demonstrate the effect of antihypertensive treatment on CHD. In paper I based on the ULSAM population the mean time to the events was at least 8 years from the start of follow-up, which also strengthens the fact that longer follow-ups are needed.

In view of an expected steep increase in the incidence of type 2 diabetes it appears important to avoid further deterioration of insulin sensitivity in al-ready insulin resistant subjects. Hence, a better screening of hypertensive patients in order to detect prevailing insulin resistance is needed to individu-alize the antihypertensive treatment.

53

Differences in clinical manifestations of coronary heart diseaseFrom a clinical point of view it is clear that CHD can vary substantially in its presentation with manifestations such as stable and unstable angina pectoris, unrecognised MI and diagnosed MI with or without ST-elevation and with or without previous signs of cardiac ischemia.

Incidence During the last decades the incidence of MI has decreased while hospitalisa-tions due to unstable angina pectoris has increased208. These changes are probably to a large extent reflecting changes in coronary risk factors, im-provements in coronary care and a possible diagnostic coding drift. Unrec-ognised myocardial infarctions have been estimated to constitute 18-35% of the total number of cases of MI11-16, but the incidence and prevalence may be underestimated since it is not unusual that typical ECG findings for MI disappear with time.

PrognosisThe prognosis associated with different manifestations of CHD has been examined in several studies. Early investigations of the prognosis associated with uncomplicated angina showed a prognosis just as bad as for MI209 210.However, more recent population-based studies of prevalent CHD have con-cluded that patients with angina have a better prognosis compared to patients with MI, but still a reduced survival compared with individuals without CHD211-213. In the case of unrecognised MI, the prognosis appears to be almost the same as for diagnosed MI11-16.

The prognostic significance of different ECG abnormalities, such as Q-wave or QS-pattern, has been extensively examined in several epidemiologi-cal studies and a strong association with cardiovascular morbidity and mor-tality has been found23-32. This was confirmed in paper III where a new Q/QS-pattern on the resting ECG was a significant predictor of cardiovascu-lar and total mortality even when the analyses were adjusted for smoking, diabetes, hypertension, hyperlipidemia and MI diagnosis over a time-period of almost 10 years.

What is the pathophysiological background to the increased mortality risk associated with the Q/QS-pattern? In paper III the Q/QS-pattern was associ-ated with impaired insulin secretion and the prevalence of diabetes was in-creased compared to a control group. However, Q/QS-pattern was a signifi-cant predictor of total and cardiovascular mortality even when the effect of diabetes was accounted for in the multiple analyses. Thus, it is not only the diabetic state per se that causes the increased mortality in these subjects. It seems rather that the Q/QS pattern indicates an objective sign of myocardial damage that may lead to heart failure or conductance abnormalities, and thus

54

increased risk of premature death. Individuals with adverse risk factor pro-files should therefore have frequent ECG monitoring and subjects with anew Q/QS-pattern on the routine ECG should be given a high priority to preventive measures against both CHD and diabetes.

The relative risks associated with ECG-findings and CHD diagnosis have been shown to attenuate over long follow-ups probably due to changes in risk factor patterns, selective mortality and interventions, such as CABG and PTCA. However, the excess mortality risk for subjects with CHD compared to those without do persist over follow-ups as long as 15 years212 213.

Risk Factors Can differences in presentation, incidence and prognosis be explained by different pathophysiology and risk factor patterns in subjects with stable and unstable CHD?

The major pathophysiological difference between acute coronary syn-dromes and stable angina pectoris is the rupture of an atherosclerotic plaque with subsequent thrombosis formation that causes acute coronary events33 35.Hypertension, smoking, hyperlipidemia and diabetes are the major abnor-malities that are generally accepted as risk factors for all manifestations of CHD2, while inflammatory cell activity, disturbed fibrinolysis and hemody-namic factors38 40 41 112 138 also have been suggested to be potential risk fac-tors for plaque rupture and thrombosis.

Unrecognised MI has been associated with increased prevalence of hyper-tension and diabetes15 17 18 in some studies, while others have failed to show differences in risk factor patterns between diagnosed and unrecognised MI2122.

Some previous studies have examined differences in risk factor profiles between subjects with stable and unstable CHD. In one study stable CHD was associated with increased levels of HDL cholesterol and increased prevalence of diabetes9, while in another study no differences between the different manifestations of CHD was found10.

In paper II serum proinsulin and blood pressure were independent risk factors for MI without previous known CHD, but not for angina pectoris demanding revascularisation with PTCA or CABG. Indeed there is a bio-logical plausibility for these findings in that elevated concentrations of pro-insulin are known to be strong determinants of PAI-1 activity, and may thereby cause impaired fibrinolysis and increased risk of thrombosis 108-110.Hypertension may promote the development of atherosclerosis by causing impaired endothelial function134-136, and may also be a trigger of acute plaque disruption by inducing mechanical stress on the arterial wall137 138.

Thus, current knowledge on the pathophysiology of plaque rupture may strengthen the hypothesis that there indeed exists a biological ground for the differences in clinical presentation. The result of paper II may implicate that treatment of insulin resistance and thus lowering proinsulin and treatment of

55

hypertension may be favourable in stabilizing the atherosclerotic plaque.However, the findings in paper II are only valid for the specific groups iden-tified in the study and cannot be transferred to general angina pectoris and MI populations. Further investigations concerning risk factor patterns asso-ciated with different clinical manifestations of CHD are therefore needed.

The metabolic syndrome and coronary heart disease; future perspectives and clinical implications How to find subjects at high risk The diagnosis of the metabolic syndrome implies a substantial additional risk of cardiovascular morbidity and mortality also above and beyond the individual risk factors86 87. However, it has been shown that proatherogenic metabolic disturbances already in the prediabetic state120 121 may contribute to the risk of macrovascular disease as much as clinical diabetes in itself. It is therefore important to identify individuals at high risk even before the development of disturbances in traditional risk markers in order to prevent the development of diabetes and CHD. New risk markers such as CRP, apolipoproteins and serum proinsulin may help to identify these individuals.

Serum proinsulin has been shown to be a strong predictor of both diabetes and CHD also in non-diabetic populations105 106 214. By using serum proinsu-lin instead of fasting blood glucose as a marker of increased risk, it may therefore be possible to capture individuals who are at high risk of develop-ing CHD although not yet being diabetic.

Recently apoB and apoA1 and especially the apoB/apoA1 ratio have emerged as powerful predictors of CHD risk123. It is plausible that these markers of dyslipidemia contain more information concerning lipid status than conventional markers, such as LDL- and HDL-cholesterol. Moreover, it has been shown that apoB and apoA1 retain their predictive power in pa-tients receiving lipid-lowering agents215. There are also methodological ad-vantages since the apolipoproteins are measured directly in blood samples123

while LDL-cholesterol most often is calculated using the Friedewald for-mula.

The methods for measuring apoB and apoA1 are internationally standard-ized216 217 and can therefore be used in clinical practice, while standardiza-tion for the analysis of serum proinsulin is yet lacking.

In paper IV proinsulin and the apoB/apoA1 ratio were included in a scor-ing scheme for the risk of fatal and nonfatal MI. This score was highly pre-dictive for future MI and showed better performance in separating those who developed MI from those who did not compared to Framingham and PRO-CAM scores, although the difference between the areas under the curves was

56

only borderline significant (p=0.08). Future investigations in larger popula-tions may confirm that these new risk markers indeed do contain more in-formation than traditional risk factors.

Treatment of the metabolic syndrome Life style changes should always be the first treatment of choice for the metabolic syndrome. Intervention trials have concluded that life style changes can reduce body weight, blood pressure and serum triglycerides, increase HDL-cholesterol and lower the risk of progression from impaired fasting glucose to type 2 diabetes by 58%218 219.

However, there most certainly exists a genetic component in the patho-physiology of insulin resistance and the metabolic syndrome, and therefore life style changes may not be enough to reduce cardiovascular risk. Treat-ment with insulin-sensitising agents, such as the thiazolidinediones, has not only been shown to increase insulin sensitivity, but also to decrease levels of CRP220, and thus these agents may have a potential to reduce cardiovascular events in subjects with insulin resistance and type 2 diabetes. Moreover, matrix metalloproteinase-9, a marker of plaque instability, has also been decreased during treatment with these agents220 and hence they may have a plaque-stabilizing effect. Ongoing clinical trials of the effects of insulin-sensitisers on cardiovascular outcomes will bring more information about the potential of these agents in treating insulin resistance and type 2 diabetes.

Since cardiovascular disease is estimated to be the leading cause of death in the world in a not too remote future, there is an ongoing development of new drugs. The so-called “dual PPAR agonists” that activate both PPAR-and PPAR- , may offer a treatment for the major metabolic disturbances of the metabolic syndrome, i.e. lower triglycerides, raise HDL-cholesterol and increase insulin sensitivity221. Another strategy to deal with high risk factors for CHD has been advocated by Wald and Law who in a recent publication suggested that aspirin, blood pressure lowering agents, a statin and folic acid should be merged together in a “polypill”. According to a meta-analysis of several studies, this treatment would prevent more than 80% of CHD events if given to everyone with existing cardiovascular disease and to everyone over the age of 55 222.

However, the safety of these kinds of drugs is still to be confirmed in large-scale studies. Furthermore, the risk of less interest in lifestyle changes as a consequence of the introduction of a universal polypharmacy with a “polypill” must be taken under consideration.

57

CONCLUSIONS

I. Study I emphasized the existence of an insulin resistant state with elevated proinsulin concentrations resulting in increased fasting blood glucose concentrations during antihypertensive treatment with beta-blockers and thiazide diuretics. Both proin-sulin concentrations and increase in fasting blood glucose were associated with increased risk of developing future MI in those on antihypertensive treatment with beta-blockers and thiazide diuretics.

II. Serum lipids were important risk factors for the development of both angina pectoris demanding revascularisation and acute MI without previous known CHD. In addition, proinsulin and blood pressure were independent predictors of MI only, suggesting these factors to be involved in thrombosis and plaque rupture.

III. The finding of a new Q/QS-pattern on the resting ECG, regard-less of history of MI, was associated with impaired insulin se-cretion and was an independent predictor of total and cardiovas-cular mortality. Therefore, subjects with a new Q/QS-pattern on the routine ECG must be given a high priority to preventive measures against both CHD and diabetes.

IV. A risk prediction score for MI including proinsulin and the ratio between apolipoprotein B and apolipoprotein A1 was developed in middle-aged men. This score was predictive for future fatal and nonfatal MI, and proved to be at least as good as the Fram-ingham and the PROCAM scores, being based on traditional risk factors.

58

ACKNOWLEDGEMENTS

I wish to express my sincere gratitude to everyone who has contributed to this work in different ways. My special thanks goes to

Lars Lind, principal supervisor, for his never-ending enthusiasm and pa-tience and for always being available for discussions and support.

Hans Lithell, previous head of the Section of Geriatrics, second supervisor, for giving me access to the ULSAM cohort and sharing his extensive knowl-edge of insulin resistance and hypertension.

Björn Zethelius, co-author, for reading my manuscripts and giving me con-structive criticism and for nice chats concerning proinsulin and life in gen-eral.

Lars Berglund, for patient and repeated statistical aid to a statistical illiter-ate…

Rawya Moshen, for providing me with datasets and helping me with my computer.

Johan Ä and Johan S, for teaching me the mysteries of JMP and Endnote.

Merike Boberg, for good advice in a tricky situation.

Previous and current colleagues and PhD students at the Section of Geriat-rics and the Section of Clinical Nutrition; Bengt, Margareta, Arvo, Kristina, Liisa, Bernice, Ann-Christin, Martin, Cecilia, Johanna, Ulf, Achraf, Anette, Agneta, Annika, Samar, Brita, Siv, Eva, Barbro for nice conversations at coffee breaks.

Karin Modin for help with all kind of practical issues.

All the brave Uppsala-men, for participating in the ULSAM study.

59

Barbro Eriksson, head of the Department of Endocrine Oncology, my boss, for allowing me free time from my clinical practice to complete this thesis, and for being a very nice person.

Katarina, for being a good friend I can always count on, and for being my sister.

Anna and Lars, my dear parents, for endless love and support, baby-sitting, cooking, restorations of apartments and houses, taming of raspberry bushes and much more.

Lars, Hanna and Jerker, all that really matters in my life.

And finally Myself

This work was supported by grants from Ernfors Fund for Diabetes Re-search, the Foundation for Geriatric Research, "Förenade Liv" Mutual Group Life Insurance Company, the Medical Faculty at Uppsala University, the Swedish Council for Planning and Coordination of Research (No. A19-5/62), the Swedish Diabetes Association, the Swedish Medical Research Council (No.5446), the Swedish National Association Against Heart and Lung Disease, and Trygg Hansa Research Fund.

60

REFERENCES

1. C.Timmreck T. An Introduction To Epidemiology, Third Edition. Jones& Bartlett publ. Sudbury, Massachusetts 2002:2-3.

2. Castelli WP. Epidemiology of coronary heart disease: the Framingham study. AmJ Med 1984;76(2A):4-12.

3. European Cardiovascular Disease statistics 2000 BHF. 4. Heart disease and stroke statistics 2003 update A. 5. Bogaty P, Brecker SJ, White SE, Stevenson RN, el-Tamimi H, Balcon R, et al.

Comparison of coronary angiographic findings in acute and chronic first presen-tation of ischemic heart disease. Circulation 1993;87(6):1938-46.

6. de Bono DP, Bhattacharrya AK. Segmental analysis of coronary arterial stenoses in patients presenting with angina or first myocardial infarction. Int J Cardiol1991;32(3):313-22.

7. Cianflone D, Ciccirillo F, Buffon A, Trani C, Scabbia EV, Finocchiaro ML, et al. Comparison of coronary angiographic narrowing in stable angina pectoris, un-stable angina pectoris, and in acute myocardial infarction. Am J Cardiol1995;76(4):215-9.

8. Crea F, Gaspardone A, Tomai F, Shoulders C, De Fazio A, Versaci F, et al. Risk factors in schoolchildren associated with a family history of unheralded myo-cardial infarction or uncomplicated stable angina in male relatives. J Am Coll Cardiol 1994;23(6):1472-8.

9. Gaspardone A, Crea F, Perino M, Iamele M, Tomai F, Versaci F, et al. Risk fac-tors in patients with different clinical and angiographic manifestations of ischemic heart disease. Cardiologia 1995;40(9):679-84.

10. Bogaty P, Robitaille NM, Solymoss S, Boyer L, Auger D, Labbe L, et al. Atherogenic, hemostatic, and other potential risk markers in subjects with pre-vious isolated myocardial infarction compared with long- standing uncompli-cated stable angina. Am Heart J 1998;136(5):884-93.

11. Kannel WB. Silent myocardial ischemia and infarction: insights from the Fram-ingham Study. Cardiol Clin 1986 Nov;4(4):583-91. 1986.

12. Kannel WB, Abbott RD. Incidence and prognosis of unrecognized myocardial infarction. An update on the Framingham study. N Engl J Med 1984 Nov 1;311(18):1144-7. 1984.

13. Bertolet BD, Hill JA. Unrecognized myocardial infarction. Cardiovasc Clin 1989;20(1):173-82. 1989.

14. Nadelmann J, Frishman WH, Ooi WL, Tepper D, Greenberg S, Guzik H, et al. Prevalence, incidence and prognosis of recognized and unrecognized myocar-dial infarction in persons aged 75 years or older: The Bronx Aging Study. Am J Cardiol 1990 Sep 1;66(5):533-7. 1990.

15. Kannel WB, Cupples LA, Gagnon DR. Incidence, precursors and prognosis of unrecognized myocardial infarction. Adv Cardiol 1990;37:202-14. 1990.

16. Sigurdsson E, Thorgeirsson G, Sigvaldason H, Sigfusson N. Unrecognized myo-cardial infarction: epidemiology, clinical characteristics, and the prognostic role

61

of angina pectoris. The Reykjavik Study. Ann Intern Med 1995 Jan 15;122(2):96-102. 1995.

17. Margolis JR, Kannel WS, Feinleib M, Dawber TR, McNamara PM. Clinical features of unrecognized myocardial infarction--silent and symptomatic. Eight-een year follow-up: the Framingham study. Am J Cardiol 1973 Jul;32(1):1-7.1973.

18. Yano K, MacLean CJ. The incidence and prognosis of unrecognized myocardial infarction in the Honolulu, Hawaii, Heart Program. Arch Intern Med 1989 Jul;149(7):1528-32. 1989.

19. Niakan E, Harati Y, Rolak LA, Comstock JP, Rokey R. Silent myocardial infarc-tion and diabetic cardiovascular autonomic neuropathy. Arch Intern Med 1986 Nov;146(11):2229-30. 1986.

20. Faerman I, Faccio E, Milei J, Nunez R, Jadzinsky M, Fox D, et al. Autonomic neuropathy and painless myocardial infarction in diabetic patients. Histologic evidence of their relationship. Diabetes 1977 Dec;26(12):1147-58. 1977.

21. Medalie JH, Goldbourt U. Unrecognized myocardial infarction: five-year inci-dence, mortality, and risk factors. Ann Intern Med 1976 May;84(5):526-31.1976.

22. Airaksinen KE, Koistinen MJ. Association between silent coronary artery dis-ease, diabetes, and autonomic neuropathy. Fact of fallacy? Diabetes Care 1992 Feb;15(2):288-92. 1992.

23. Rose G, Baxter PJ, Reid DD, McCartney P. Prevalence and prognosis of electro-cardiographic findings in middle-aged men. Br Heart J 1978 Jun;40(6):636-43.1978.

24. Cedres BL, Liu K, Stamler J, Dyer AR, Stamler R, Berkson DM, et al. Inde-pendent contribution of electrocardiographic abnormalities to risk of death from coronary heart disease, cardiovascular diseases and all causes. Findings of three Chicago epidemiologic studies. Circulation 1982 Jan;65(1):146-53. 1982.

25. Cullen K, Stenhouse NS, Wearne KL, Cumpston GN. Electrocardiograms and 13 year cardiovascular mortality in Busselton study. Br Heart J 1982 Mar;47(3):209-12. 1982.

26. Kannel WB, Anderson K, McGee DL, Degatano LS, Stampfer MJ. Nonspecific electrocardiographic abnormality as a predictor of coronary heart disease: the Framingham Study. Am Heart J 1987;113(2 Pt 1):370-6.

27. Liao YL, Liu KA, Dyer A, Schoenberger JA, Shekelle RB, Colette P, et al. Ma-jor and minor electrocardiographic abnormalities and risk of death from coro-nary heart disease, cardiovascular diseases and all causes in men and women. JAm Coll Cardiol 1988;12(6):1494-500.

28. Knutsen R, Knutsen SF, Curb JD, Reed DM, Kautz JA, Yano K. The predictive value of resting electrocardiograms for 12-year incidence of coronary heart dis-ease in the Honolulu Heart Program. J Clin Epidemiol 1988;41(3):293-302.

29. Tervahauta M, Pekkanen J, Punsar S, Nissinen A. Resting electrocardiographic abnormalities as predictors of coronary events and total mortality among elderly men. Am J Med 1996 Jun;100(6):641-5. 1996.

30. De Bacquer D, De Backer G, Kornitzer M, Blackburn H. Prognostic value of ECG findings for total, cardiovascular disease, and coronary heart disease death in men and women. Heart 1998 Dec;80(6):570-7. 1998.

31. Crow RS, Prineas RJ, Hannan PJ, Grandits G, Blackburn H. Prognostic associa-tions of Minnesota Code serial electrocardiographic change classification with coronary heart disease mortality in the Multiple Risk Factor Intervention Trial. Am J Cardiol 1997 Jul 15;80(2):138-44. 1997.

62

32. Menotti A, Mulder I, Kromhout D, Nissinen A, Feskens EJ, Giampaoli S. The association of silent electrocardiographic findings with coronary deaths among elderly men in three European countries. The FINE study. Acta Cardiol 2001 Feb;56(1):27-36. 2001.

33. Davies MJ, Thomas AC. Plaque fissuring--the cause of acute myocardial infarc-tion, sudden ischaemic death, and crescendo angina. Br Heart J 1985;53(4):363-73.

34. Fuster V, Badimon L, Badimon JJ, Chesebro JH. The pathogenesis of coronary artery disease and the acute coronary syndromes (2). N Engl J Med1992;326(5):310-8.

35. Fuster V, Badimon L, Badimon JJ, Chesebro JH. The pathogenesis of coronary artery disease and the acute coronary syndromes (1). N Engl J Med1992;326(4):242-50.

36. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ. Compensa-tory enlargement of human atherosclerotic coronary arteries. N Engl J Med1987;316(22):1371-5.

37. Mann JM, Davies MJ. Vulnerable plaque. Relation of characteristics to degree of stenosis in human coronary arteries. Circulation 1996;94(5):928-31.

38. Falk E, Shah PK, Fuster V. Coronary plaque disruption. Circulation1995;92(3):657-71.

39. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med1999;340(2):115-26.

40. Moreno PR, Falk E, Palacios IF, Newell JB, Fuster V, Fallon JT. Macrophage infiltration in acute coronary syndromes. Implications for plaque rupture. Circu-lation 1994;90(2):775-8.

41. Shah PK, Falk E, Badimon JJ, Fernandez-Ortiz A, Mailhac A, Villareal-Levy G, et al. Human monocyte-derived macrophages induce collagen breakdown in fi-brous caps of atherosclerotic plaques. Potential role of matrix-degrading metal-loproteinases and implications for plaque rupture. Circulation 1995;92(6):1565-9.

42. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflamma-tion, aspirin, and the risk of cardiovascular disease in apparently healthy men. NEngl J Med 1997;336(14):973-9.

43. Haverkate F, Thompson SG, Pyke SD, Gallimore JR, Pepys MB. Production of C-reactive protein and risk of coronary events in stable and unstable angina. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group. Lancet 1997;349(9050):462-6.

44. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease. Diabetes 1988;37(12):1595-607.

45. Alberti KG, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 1998 Jul;15(7):539-53. 1998.

46. Balkau B, Charles MA. Comment on the provisional report from the WHO con-sultation. European Group for the Study of Insulin Resistance (EGIR). Diabet Med 1999;16(5):442-3.

47. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001;285(19):2486-97.

48. Balkau B, Charles MA, Drivsholm T, Borch-Johnsen K, Wareham N, Yudkin JS, et al. Frequency of the WHO metabolic syndrome in European cohorts, and

63

an alternative definition of an insulin resistance syndrome. Diabetes Metab2002;28(5):364-76.

49. Scott CL. Diagnosis, prevention, and intervention for the metabolic syndrome. Am J Cardiol 2003;92(1A):35i-42i.

50. Reaven GM. The Pathophysiological Consequences of Adipose Tissue Insulin resistance. In: Reaven G, Laws A, editors. Insulin Resistance, The Metabolic syndrome X 1999;Totowa:Humana Press Inc:233-246.

51. Radzuik J PS. The Role of the Liver in Insulin Action and Resistance. In: Reaven G, Laws A, editors. Insulin Resistance, The Metabolic syndrome X1999;Totowa:Humana Press Inc:197-231.

52. Hollenbeck C, Reaven GM. Variations in insulin-stimulated glucose uptake in healthy individuals with normal glucose tolerance. J Clin Endocrinol Metab1987;64(6):1169-73.

53. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol 1979 Sep;237(3):E214-23. 1979.

54. Wang PW, Abbasi F, Carantoni M, Chen YD, Azhar S, Reaven GM. Insulin resistance does not change the ratio of proinsulin to insulin in normal volun-teers. J Clin Endocrinol Metab 1997;82(10):3221-4.

55. Phillips DI, Clark PM, Hales CN, Osmond C. Understanding oral glucose toler-ance: comparison of glucose or insulin measurements during the oral glucose tolerance test with specific measurements of insulin resistance and insulin secre-tion. Diabet Med 1994;11(3):286-92.

56. Depres J MA. Obesity and Insulin Resistance. In: Reaven G, Laws A, editors. Insulin Resistance, The Metabolic syndrome X 1999;Totowa:Humana Press Inc:51-82.

57. Dela F MK, Galbo H. Physical Activity and Insulin Resistance in Man. In: Reaven G, Laws A, editors. Insulin Resistance, The Metabolic syndrome X1999;Totowa:Humana Press Inc:97-120.

58. Eliasson S. Insulin Resistance in Smokers. In: Reaven G, Laws A, editors. Insu-lin Resistance, The Metabolic syndrome X 1999;Totowa:Humana Press Inc:121-136.

59. Riserus U, Basu S, Jovinge S, Fredrikson GN, Arnlov J, Vessby B. Supplemen-tation with conjugated linoleic acid causes isomer-dependent oxidative stress and elevated C-reactive protein: a potential link to fatty acid-induced insulin re-sistance. Circulation 2002;106(15):1925-9.

60. Carantoni M, Abbasi F, Warmerdam F, Klebanov M, Wang PW, Chen YD, et al. Relationship between insulin resistance and partially oxidized LDL particles in healthy, nondiabetic volunteers. Arterioscler Thromb Vasc Biol 1998;18(5):762-7.

61. Gopaul NK, Manraj MD, Hebe A, Lee Kwai Yan S, Johnston A, Carrier MJ, et al. Oxidative stress could precede endothelial dysfunction and insulin resistance in Indian Mauritians with impaired glucose metabolism. Diabetologia2001;44(6):706-12.

62. Arnlov J, Vessby B, Riserus U. Coffee consumption and insulin sensitivity. JAMA 2004;291(10):1199-201.

63. Lithell HO, McKeigue PM, Berglund L, Mohsen R, Lithell UB, Leon DA. Rela-tion of size at birth to non-insulin dependent diabetes and insulin concentrations in men aged 50-60 years. BMJ 1996;312(7028):406-10.

64. Phillips DI, Barker DJ, Hales CN, Hirst S, Osmond C. Thinness at birth and insulin resistance in adult life. Diabetologia 1994;37(2):150-4.

64

65. Byberg L, McKeigue PM, Zethelius B, Lithell HO. Birth weight and the insulin resistance syndrome: association of low birth weight with truncal obesity and raised plasminogen activator inhibitor-1 but not with abdominal obesity or plasma lipid disturbances. Diabetologia 2000;43(1):54-60.

66. Andrews RC, Walker BR. Glucocorticoids and insulin resistance: old hormones, new targets. Clin Sci (Lond) 1999;96(5):513-23.

67. Reynolds RM, Chapman KE, Seckl JR, Walker BR, McKeigue PM, Lithell HO. Skeletal muscle glucocorticoid receptor density and insulin resistance. JAMA2002;287(19):2505-6.

68. Ford ES. The metabolic syndrome and C-reactive protein, fibrinogen, and leuko-cyte count: findings from the Third National Health and Nutrition Examination Survey. Atherosclerosis 2003;168(2):351-8.

69. Festa A, D'Agostino R, Jr., Howard G, Mykkanen L, Tracy RP, Haffner SM. Chronic subclinical inflammation as part of the insulin resistance syndrome: the Insulin Resistance Atherosclerosis Study (IRAS). Circulation 2000;102(1):42-7.

70. Festa A, Hanley AJ, Tracy RP, D'Agostino R, Jr., Haffner SM. Inflammation in the prediabetic state is related to increased insulin resistance rather than de-creased insulin secretion. Circulation 2003;108(15):1822-30.

71. Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA2001;286(3):327-34.

72. Pickup JC, Crook MA. Is type II diabetes mellitus a disease of the innate im-mune system? Diabetologia 1998;41(10):1241-8.

73. Julius S, Jamerson K. Sympathetics, insulin resistance and coronary risk in hy-pertension: the 'chicken-and-egg' question. J Hypertens 1994;12(5):495-502.

74. Lithell H, Lindgarde F, Hellsing K, Lundqvist G, Nygaard E, Vessby B, et al. Body weight, skeletal muscle morphology, and enzyme activities in relation to fasting serum insulin concentration and glucose tolerance in 48-year-old men. Diabetes 1981;30(1):19-25.

75. Lillioja S, Young AA, Culter CL, Ivy JL, Abbott WG, Zawadzki JK, et al. Skeletal muscle capillary density and fiber type are possible determinants of in vivo insulin resistance in man. J Clin Invest 1987;80(2):415-24.

76. Marin P, Andersson B, Krotkiewski M, Bjorntorp P. Muscle fiber composition and capillary density in women and men with NIDDM. Diabetes Care1994;17(5):382-6.

77. Haffner SM, Stern MP, Hazuda HP, Mitchell BD, Patterson JK. Increased insu-lin concentrations in nondiabetic offspring of diabetic parents. N Engl J Med1988;319(20):1297-301.

78. Mandarino. Skeletal Muscle Insulin Resistance in Humans; Cellular Mecha-nisms. In: Reaven G, Laws A, editors. Insulin Resistance, The Metabolic syn-drome X 1999;Totowa:Humana Press Inc:179-196.

79. Tooke J. The association between insulin resistance and endotheliopathy. Diabe-tes Obes Metab 1999;1(Suppl 1):S17-22.

80. Storlien LH, Pan DA, Kriketos AD, O'Connor J, Caterson ID, Cooney GJ, et al. Skeletal muscle membrane lipids and insulin resistance. Lipids1996;31(Suppl):S261-5.

81. Vessby B, Tengblad S, Lithell H. Insulin sensitivity is related to the fatty acid composition of serum lipids and skeletal muscle phospholipids in 70-year-old men. Diabetologia 1994;37(10):1044-50.

82. Vessby B. Dietary fat and insulin action in humans. Br J Nutr 2000;83(Suppl 1):S91-6.

65

83. Petersen KF, Dufour S, Befroy D, Garcia R, Shulman GI. Impaired mitochon-drial activity in the insulin-resistant offspring of patients with type 2 diabetes. NEngl J Med 2004;350(7):664-71.

84. Banerji MA, Buckley MC, Chaiken RL, Gordon D, Lebovitz HE, Kral JG. Liver fat, serum triglycerides and visceral adipose tissue in insulin-sensitive and insu-lin-resistant black men with NIDDM. Int J Obes Relat Metab Disord1995;19(12):846-50.

85. Goto T, Onuma T, Takebe K, Kral JG. The influence of fatty liver on insulin clearance and insulin resistance in non-diabetic Japanese subjects. Int J Obes Relat Metab Disord 1995;19(12):841-5.

86. Isomaa B, Almgren P, Tuomi T, Forsen B, Lahti K, Nissen M, et al. Cardiovas-cular morbidity and mortality associated with the metabolic syndrome. Diabetes Care 2001 Apr;24(4):683-9. 2001.

87. Lakka HM, Laaksonen DE, Lakka TA, Niskanen LK, Kumpusalo E, Tuomilehto J, et al. The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA 2002;288(21):2709-16.

88. Fuller JH, Shipley MJ, Rose G, Jarrett RJ, Keen H. Coronary-heart-disease risk and impaired glucose tolerance. The Whitehall study. Lancet1980;1(8183):1373-6.

89. Andersson OK, Almgren T, Persson B, Samuelsson O, Hedner T, Wilhelmsen L. Survival in treated hypertension: follow up study after two decades. BMJ1998;317(7152):167-71.

90. Assmann G, Schulte H. Relation of high-density lipoprotein cholesterol and triglycerides to incidence of atherosclerotic coronary artery disease (the PRO-CAM experience). Prospective Cardiovascular Munster study. Am J Cardiol1992;70(7):733-7.

91. Borch-Johnsen K, Feldt-Rasmussen B, Strandgaard S, Schroll M, Jensen JS. Urinary albumin excretion. An independent predictor of ischemic heart disease. Arterioscler Thromb Vasc Biol 1999;19(8):1992-7.

92. Rosengren A, Wedel H, Wilhelmsen L. Body weight and weight gain during adult life in men in relation to coronary heart disease and mortality. A prospec-tive population study. Eur Heart J 1999;20(4):269-77.

93. Ruige JB, Assendelft WJ, Dekker JM, Kostense PJ, Heine RJ, Bouter LM. Insu-lin and risk of cardiovascular disease: a meta-analysis. Circulation1998;97(10):996-1001.

94. Laakso M. How good a marker is insulin level for insulin resistance? Am J Epi-demiol 1993;137(9):959-65.

95. McKeigue P, Davey G. Associations between insulin levels and cardiovascular disease are confounded by comorbidity. Diabetes Care 1995;18(9):1294-8.

96. Bressler P, Bailey SR, Matsuda M, DeFronzo RA. Insulin resistance and coro-nary artery disease. Diabetologia 1996;39(11):1345-50.

97. Zethelius B HN, Berne C. Epidemiology of Insulin Resistance, Complications and Mortality in Type 2 Diabetes Mellitus (Abstract). Diabetologia2003;46(Suppl.2, August 2003):A 76.

98. Petrie JR, Ueda S, Webb DJ, Elliott HL, Connell JM. Endothelial nitric oxide production and insulin sensitivity. A physiological link with implications for pathogenesis of cardiovascular disease. Circulation 1996;93(7):1331-3.

99. Stout RW. The impact of insulin upon atherosclerosis. Horm Metab Res1994;26(3):125-8.

100. Stern MP. Diabetes and cardiovascular disease. The "common soil" hypothesis. Diabetes 1995;44(4):369-74.

66

101. Nagi DK, Knowler WC, Mohamed-Ali V, Bennett PH, Yudkin JS. Intact proin-sulin, des 31,32 proinsulin, and specific insulin concentrations among nondia-betic and diabetic subjects in populations at varying risk of type 2 diabetes. Diabetes Care 1998;21(1):127-33.

102. Saad MF, Kahn SE, Nelson RG, Pettitt DJ, Knowler WC, Schwartz MW, et al. Disproportionately elevated proinsulin in Pima Indians with noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1990;70(5):1247-53.

103. Revers RR, Henry R, Schmeiser L, Kolterman O, Cohen R, Bergenstal R, et al. The effects of biosynthetic human proinsulin on carbohydrate metabolism. Dia-betes 1984;33(8):762-70.

104. Glauber HS, Revers RR, Henry R, Schmeiser L, Wallace P, Kolterman OG, et al. In vivo deactivation of proinsulin action on glucose disposal and hepatic glu-cose production in normal man. Diabetes 1986;35(3):311-7.

105. Yudkin JS, May M, Elwood P, Yarnell JW, Greenwood R, Davey Smith G. Concentrations of proinsulin like molecules predict coronary heart disease risk independently of insulin: prospective data from the Caerphilly Study. Diabe-tologia 2002;45(3):327-36.

106. Zethelius B, Byberg L, Hales CN, Lithell H, Berne C. Proinsulin is an inde-pendent predictor of coronary heart disease: Report from a 27-year follow-up study. Circulation 2002 May 7;105(18):2153-8. 2002.

107. Galloway JA, Hooper SA, Spradlin CT, Howey DC, Frank BH, Bowsher RR, et al. Biosynthetic human proinsulin. Review of chemistry, in vitro and in vivo receptor binding, animal and human pharmacology studies, and clinical trial ex-perience. Diabetes Care 1992;15(5):666-92.

108. Gray RP, Panahloo A, Mohamed-Ali V, Patterson DL, Yudkin JS. Proinsulin-like molecules and plasminogen activator inhibitor type 1 (PAI-1) activity in diabetic and non-diabetic subjects with and without myocardial infarction. Atherosclerosis 1997;130(1-2):171-8.

109. Nordt TK, Sawa H, Fujii S, Sobel BE. Induction of plasminogen activator in-hibitor type-1 (PAI-1) by proinsulin and insulin in vivo. Circulation1995;91(3):764-70.

110. Schneider DJ, Nordt TK, Sobel BE. Stimulation by proinsulin of expression of plasminogen activator inhibitor type-I in endothelial cells. Diabetes1992;41(7):890-5.

111. Sobel BE, Woodcock-Mitchell J, Schneider DJ, Holt RE, Marutsuka K, Gold H. Increased plasminogen activator inhibitor type 1 in coronary artery atherec-tomy specimens from type 2 diabetic compared with nondiabetic patients: a po-tential factor predisposing to thrombosis and its persistence. Circulation1998;97(22):2213-21.

112. Sobel BE. The potential influence of insulin and plasminogen activator inhibi-tor type 1 on the formation of vulnerable atherosclerotic plaques associated with type 2 diabetes. Proc Assoc Am Physicians 1999;111(4):313-8.

113. Byberg L, Siegbahn A, Berglund L, McKeigue P, Reneland R, Lithell H. Plas-minogen activator inhibitor-1 activity is independently related to both insulin sensitivity and serum triglycerides in 70-year-old men. Arterioscler Thromb Vasc Biol 1998;18(2):258-64.

114. Klein R. Hyperglycemia and microvascular and macrovascular disease in dia-betes. Diabetes Care 1995;18(2):258-68.

115. Yki-Jarvinen H. Toxicity of hyperglycaemia in type 2 diabetes. Diabetes Metab Rev 1998;14(Suppl 1):S45-50.

67

116. Sarabi M, Millgard J, Lind L. Effects of age, gender and metabolic factors on endothelium-dependent vasodilation: a population-based study. J Intern Med1999;246(3):265-74.

117. Williams SB, Goldfine AB, Timimi FK, Ting HH, Roddy MA, Simonson DC, et al. Acute hyperglycemia attenuates endothelium-dependent vasodilation in humans in vivo. Circulation 1998;97(17):1695-701.

118. UKPDS. Intensive blood-glucose control with sulphonylureas or insulin com-pared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet 1998;352(9131):837-53.

119. Glucose tolerance and mortality: comparison of WHO and American Diabetes Association diagnostic criteria. The DECODE study group. European Diabetes Epidemiology Group. Diabetes Epidemiology: Collaborative analysis Of Diag-nostic criteria in Europe. Lancet 1999;354(9179):617-21.

120. Haffner SM, Stern MP, Hazuda HP, Mitchell BD, Patterson JK. Cardiovascular risk factors in confirmed prediabetic individuals. Does the clock for coronary heart disease start ticking before the onset of clinical diabetes? JAMA 1990;263(21):2893-8.

121. Haffner SM. Insulin resistance, inflammation, and the prediabetic state. Am J Cardiol 2003;92(4A):18J-26J.

122. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet1994;344(8934):1383-9.

123. Walldius G, Jungner I, Holme I, Aastveit AH, Kolar W, Steiner E. High apoli-poprotein B, low apolipoprotein A-I, and improvement in the prediction of fatal myocardial infarction (AMORIS study): a prospective study. Lancet2001;358(9298):2026-33.

124. Sacks FM, Alaupovic P, Moye LA, Cole TG, Sussex B, Stampfer MJ, et al. VLDL, apolipoproteins B, CIII, and E, and risk of recurrent coronary events in the Cholesterol and Recurrent Events (CARE) trial. Circulation2000;102(16):1886-92.

125. Jackson RL, Gotto AM, Stein O, Stein Y. A comparative study on the removal of cellular lipids from Landschutz ascites cells by human plasma apolipopro-teins. J Biol Chem 1975;250(18):7204-9.

126. Francis MC, Frohlich JJ. Coronary artery disease in patients at low risk--apolipoprotein AI as an independent risk factor. Atherosclerosis2001;155(1):165-70.

127. Nissen SE, Tsunoda T, Tuzcu EM, Schoenhagen P, Cooper CJ, Yasin M, et al. Effect of recombinant ApoA-I Milano on coronary atherosclerosis in patients with acute coronary syndromes: a randomized controlled trial. JAMA2003;290(17):2292-300.

128. Festa A, D'Agostino R, Jr., Mykkanen L, Tracy RP, Hales CN, Howard BV, et al. LDL particle size in relation to insulin, proinsulin, and insulin sensitivity. The Insulin Resistance Atherosclerosis Study. Diabetes Care 1999;22(10):1688-93.

129. Lamarche B, Tchernof A, Moorjani S, Cantin B, Dagenais GR, Lupien PJ, et al. Small, dense low-density lipoprotein particles as a predictor of the risk of ischemic heart disease in men. Prospective results from the Quebec Cardiovas-cular Study. Circulation 1997;95(1):69-75.

130. Ohrvall M, Berglund L, Salminen I, Lithell H, Aro A, Vessby B. The serum cholesterol ester fatty acid composition but not the serum concentration of alpha

68

tocopherol predicts the development of myocardial infarction in 50-year-old men: 19 years follow-up. Atherosclerosis 1996;127(1):65-71.

131. Messner T, Sihm H, Vessby B. Fatty acid composition in serum among males 4-16 years after myocardial infarction. Int J Circumpolar Health 1998;57(1):22-31.

132. Moers A, Schrezenmeir J. Palmitic acid but not stearic acid inhibits NO-production in endothelial cells. Exp Clin Endocrinol Diabetes 1997;105(Suppl 2):78-80.

133. Kannel WB. Role of blood pressure in cardiovascular disease: the Framingham Study. Angiology 1975;26(1 Pt. 1):1-14.

134. Panza JA, Quyyumi AA, Brush JE, Jr., Epstein SE. Abnormal endothelium-dependent vascular relaxation in patients with essential hypertension. N Engl J Med 1990;323(1):22-7.

135. 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(6):1762-7.

136. Lind L, Granstam SO, Millgard J. Endothelium-dependent vasodilation in hy-pertension: a review. Blood Press 2000;9(1):4-15.

137. Lee RT, Grodzinsky AJ, Frank EH, Kamm RD, Schoen FJ. Structure-dependent dynamic mechanical behavior of fibrous caps from human athero-sclerotic plaques. Circulation 1991;83(5):1764-70.

138. Shah PK. Pathophysiology of plaque rupture and the concept of plaque stabili-zation. Cardiol Clin 1996;14(1):17-29.

139. Sjogren LS, Doroudi R, Gan L, Jungersten L, Hrafnkelsdottir T, Jern S. Ele-vated intraluminal pressure inhibits vascular tissue plasminogen activator secre-tion and downregulates its gene expression. Hypertension 2000;35(4):1002-8.

140. Ferrannini E, Buzzigoli G, Bonadonna R, Giorico MA, Oleggini M, Graziadei L, et al. Insulin resistance in essential hypertension. N Engl J Med1987;317(6):350-7.

141. Pollare T, Lithell H, Berne C. Insulin resistance is a characteristic feature of primary hypertension independent of obesity. Metabolism 1990;39(2):167-74.

142. Reaven GM. Relationship between insulin resistance and hypertension. Diabe-tes Care 1991;14 Suppl 4:33-8.

143. DeFronzo RA. The effect of insulin on renal sodium metabolism. A review with clinical implications. Diabetologia 1981;21(3):165-71.

144. Rowe JW, Young JB, Minaker KL, Stevens AL, Pallotta J, Landsberg L. Effect of insulin and glucose infusions on sympathetic nervous system activity in nor-mal man. Diabetes 1981;30(3):219-25.

145. Reaven GM, Lithell H, Landsberg L. Hypertension and associated metabolic abnormalities--the role of insulin resistance and the sympathoadrenal system. NEngl J Med 1996;334(6):374-81.

146. Pollare T, Lithell H, Selinus I, Berne C. Sensitivity to insulin during treatment with atenolol and metoprolol: a randomised, double blind study of effects on carbohydrate and lipoprotein metabolism in hypertensive patients. BMJ1989;298(6681):1152-7.

147. Pollare T, Lithell H, Berne C. A comparison of the effects of hydrochlorothi-azide and captopril on glucose and lipid metabolism in patients with hyperten-sion. N Engl J Med 1989;321(13):868-73.

148. Lindholm LH, Persson M, Alaupovic P, Carlberg B, Svensson A, Samuelsson O. Metabolic outcome during 1 year in newly detected hypertensives: results of the Antihypertensive Treatment and Lipid Profile in a North of Sweden Efficacy Evaluation (ALPINE study). J Hypertens 2003;21(8):1563-74.

69

149. Lundgren H, Bjorkman L, Keiding P, Lundmark S, Bengtsson C. Diabetes in patients with hypertension receiving pharmacological treatment. BMJ1988;297(6662):1512.

150. Skarfors ET, Selinus KI, Lithell HO. Risk factors for developing non-insulin dependent diabetes: a 10 year follow up of men in Uppsala. BMJ1991;303(6805):755-60.

151. Mykkanen L, Kuusisto J, Pyorala K, Laakso M, Haffner SM. Increased risk of non-insulin-dependent diabetes mellitus in elderly hypertensive subjects. J Hy-pertens 1994;12(12):1425-32.

152. Gress TW, Nieto FJ, Shahar E, Wofford MR, Brancati FL. Hypertension and antihypertensive therapy as risk factors for type 2 diabetes mellitus. Atheroscle-rosis Risk in Communities Study. N Engl J Med 2000;342(13):905-12.

153. Lithell H, Pollare T, Vessby B. Metabolic effects of pindolol and propranolol in a double-blind cross-over study in hypertensive patients. Blood Press1992;1(2):92-101.

154. Lund-Johansen P, Omvik P, Nordrehaug JE. Long-term hemodynamic effects of antihypertensive treatment. Clin Investig 1992;70(Suppl 1):S58-64.

155. Berne C, Pollare T, Lithell H. Effects of antihypertensive treatment on insulin sensitivity with special reference to ACE inhibitors. Diabetes Care1991;14(Suppl 4):39-47.

156. Pollare T, Lithell H, Selinus I, Berne C. Application of prazosin is associated with an increase of insulin sensitivity in obese patients with hypertension. Dia-betologia 1988;31(7):415-20.

157. Haenni A, Lithell H. Treatment with a beta-blocker with beta 2-agonism im-proves glucose and lipid metabolism in essential hypertension. Metabolism1994;43(4):455-61.

158. Swislocki AL. Impaired insulin clearance in essential hypertension. J Hum Hypertens 1994;8(3):185-90.

159. Rossner S, Taylor CL, Byington RP, Furberg CD. Long term propranolol treatment and changes in body weight after myocardial infarction. BMJ1990;300(6729):902-3.

160. Andersson OK, Gudbrandsson T, Jamerson K. Metabolic adverse effects of thiazide diuretics: the importance of normokalaemia. J Intern Med Suppl1991;735:89-96.

161. Lake CR, Ziegler MG, Coleman MD, Kopin IJ. Hydrochlorothiazide-induced sympathetic hyperactivity in hypertensive patients. Clin Pharmacol Ther1979;26(4):428-32.

162. Samuelsson O, Pennert K, Andersson O, Berglund G, Hedner T, Persson B, et al. Diabetes mellitus and raised serum triglyceride concentration in treated hy-pertension--are they of prognostic importance? Observational study. BMJ1996;313(7058):660-3.

163. Hansson L, Lindholm LH, Niskanen L, Lanke J, Hedner T, Niklason A, et al. Effect of angiotensin-converting-enzyme inhibition compared with conventional therapy on cardiovascular morbidity and mortality in hypertension: the Capto-pril Prevention Project (CAPPP) randomised trial. Lancet 1999;353(9153):611-6.

164. Dahlof B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, Faire U, et al. Car-diovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet 2002;359(9311):995-1003.

165. Major outcomes in moderately hypercholesterolemic, hypertensive patients randomized to pravastatin vs usual care: The Antihypertensive and Lipid-

70

Lowering Treatment to Prevent Heart Attack Trial (ALLHAT-LLT). JAMA 2002;288(23):2998-3007.

166. Efficacy of atenolol and captopril in reducing risk of macrovascular and mi-crovascular complications in type 2 diabetes: UKPDS 39. UK Prospective Dia-betes Study Group. BMJ 1998;317(7160):713-20.

167. Wilson PW, Castelli WP, Kannel WB. Coronary risk prediction in adults (the Framingham Heart Study). Am J Cardiol 1987;59(14):91G-94G.

168. Gordon T, Kannel WB. Multiple risk functions for predicting coronary heart disease: the concept, accuracy, and application. Am Heart J 1982;103(6):1031-9.

169. Anderson KM, Wilson PW, Odell PM, Kannel WB. An updated coronary risk profile. A statement for health professionals. Circulation 1991;83(1):356-62.

170. Menotti A, Puddu PE, Lanti M. Comparison of the Framingham risk function-based coronary chart with risk function from an Italian population study. Eur Heart J 2000;21(5):365-70.

171. Thomsen TF, McGee D, Davidsen M, Jorgensen T. A cross-validation of risk-scores for coronary heart disease mortality based on data from the Glostrup Population Studies and Framingham Heart Study. Int J Epidemiol2002;31(4):817-22.

172. Assmann G, Cullen P, Schulte H. Simple scoring scheme for calculating the risk of acute coronary events based on the 10-year follow-up of the prospective cardiovascular Munster (PROCAM) study.PG - 310-5. Circulation2002;105(3):310-5.

173. Conroy RM, Pyorala K, Fitzgerald AP, Sans S, Menotti A, De Backer G, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003;24(11):987-1003.

174. Sobey WJ, Beer SF, Carrington CA, Clark PM, Frank BH, Gray IP, et al. Sen-sitive and specific two-site immunoradiometric assays for human insulin, proin-sulin, 65-66 split and 32-33 split proinsulins. Biochem J 1989;260(2):535-41.

175. Stoffel W CF, Ahrens EH. Analysis of long-chain fatty acids by gas-liquid chromatography. Anal Chem 1959;31:307-308.

176. Milne DB, Botnen J. Retinol, alpha-tocopherol, lycopene, and alpha- and beta-carotene simultaneously determined in plasma by isocratic liquid chromatogra-phy. Clin Chem 1986;32(5):874-6.

177. Blackburn, A. K, al. e. The Electrocardiogram in Population Studies. A Classi-fication System. Circulation 1960;XXI(June):1160-1175.

178. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. National Diabetes Data Group. Diabetes 1979;28(12):1039-57.

179. Hosmer DW LS. Applied Logistic Regression, second edition. New York:Wiley2000:147-156.

180. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver oper-ating characteristic (ROC) curve. Radiology 1982;143(1):29-36.

181. Hanley JA, McNeil BJ. A method of comparing the areas under receiver oper-ating characteristic curves derived from the same cases. Radiology1983;148(3):839-43.

182. Rosengren A, Wilhelmsen L, Berglund G, Elmfeldt D. Non-participants in a general population study of men, with special reference to social and alcoholic problems. Acta Med Scand 1987;221(3):243-51.

183. Värdering av diagnoskvalitet för akut hjärtinfarkt i patientregistret 1987 och 1995. Epidemiologiskt centrum, socialstyrelsen 2000.

184. Rosengren A, Eriksson H, Larsson B, Svardsudd K, Tibblin G, Welin L, et al. Secular changes in cardiovascular risk factors over 30 years in Swedish men

71

aged 50: the study of men born in 1913, 1923, 1933 and 1943. J Intern Med2000;247(1):111-8.

185. Pekkanen J, Nissinen A, Puska P, Punsar S, Karvonen MJ. Risk factors and 25 year risk of coronary heart disease in a male population with a high incidence of the disease: the Finnish cohorts of the seven countries study. BMJ1989;299(6691):81-5.

186. Pekkanen J, Tervahauta M, Nissinen A, Karvonen MJ. Does the predictive value of baseline coronary risk factors change over a 30-year follow-up? Cardi-ology 1993;82(2-3):181-90.

187. Welin L, Eriksson H, Larsson B, Svardsudd K, Wilhelmsen L, Tibblin G. Risk factors for coronary heart disease during 25 years of follow-up. The study of men born in 1913. Cardiology 1993;82(2-3):223-8.

188. Wannamethee SG, Shaper AG, Whincup PH, Walker M. Role of risk factors for major coronary heart disease events with increasing length of follow up. Heart 1999;81(4):374-9.

189. Haffner SM. Epidemiology of insulin resistance and its relation to coronary artery disease. Am J Cardiol 1999;84(1A):11J-14J.

190. Williams SB, Cusco JA, Roddy MA, Johnstone MT, Creager MA. Impaired nitric oxide-mediated vasodilation in patients with non- insulin-dependent dia-betes mellitus. J Am Coll Cardiol 1996;27(3):567-74.

191. McVeigh GE, Brennan GM, Johnston GD, McDermott BJ, McGrath LT, Henry WR, et al. Impaired endothelium-dependent and independent vasodilation in pa-tients with type 2 (non-insulin-dependent) diabetes mellitus. Diabetologia1992;35(8):771-6.

192. Potter van Loon BJ, Kluft C, Radder JK, Blankenstein MA, Meinders AE. The cardiovascular risk factor plasminogen activator inhibitor type 1 is related to in-sulin resistance. Metabolism 1993;42(8):945-9.

193. Weyer C, Bogardus C, Mott DM, Pratley RE. The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabe-tes mellitus. J Clin Invest 1999;104(6):787-94.

194. Pratley RE, Weyer C. The role of impaired early insulin secretion in the patho-genesis of Type II diabetes mellitus. Diabetologia 2001;44(8):929-45.

195. Wilhelmsen L. Coronary heart disease: epidemiology of smoking and interven-tion studies of smoking. Am Heart J 1988;115(1 Pt 2):242-9.

196. Rasky E, Stronegger WJ, Freidl W. The relationship between body weight and patterns of smoking in women and men. Int J Epidemiol 1996;25(6):1208-12.

197. Molarius A, Seidell JC, Kuulasmaa K, Dobson AJ, Sans S. Smoking and rela-tive body weight: an international perspective from the WHO MONICA Project. J Epidemiol Community Health 1997;51(3):252-60.

198. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric ap-proach. Biometrics 1988;44(3):837-45.

199. Schaefer EJ, Lamon-Fava S, Cohn SD, Schaefer MM, Ordovas JM, Castelli WP, et al. Effects of age, gender, and menopausal status on plasma low density lipoprotein cholesterol and apolipoprotein B levels in the Framingham Offspring Study. J Lipid Res 1994;35(5):779-92.

200. Garfagnini A, Devoto G, Rosselli P, Boggiano P, Venturini M. Relationship between HDL-cholesterol and apolipoprotein A1 and the severity of coronary artery disease. Eur Heart J 1995;16(4):465-70.

201. Lamarche B, Moorjani S, Lupien PJ, Cantin B, Bernard PM, Dagenais GR, et al. Apolipoprotein A-I and B levels and the risk of ischemic heart disease during

72

a five-year follow-up of men in the Quebec cardiovascular study. Circulation1996;94(3):273-8.

202. Collins R, Peto R, MacMahon S, Hebert P, Fiebach NH, Eberlein KA, et al. 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(8693):827-38.

203. Balkau B, Bertrais S, Ducimetiere P, Eschwege E. Is there a glycemic threshold for mortality risk? Diabetes Care 1999;22(5):696-9.

204. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investiga-tors. N Engl J Med 2000;342(3):145-53.

205. Wing LM, Reid CM, Ryan P, Beilin LJ, Brown MA, Jennings GL, et al. A comparison of outcomes with angiotensin-converting--enzyme inhibitors and diuretics for hypertension in the elderly. N Engl J Med 2003;348(7):583-92.

206. Mortality rates after 10.5 years for participants in the Multiple Risk Factor Intervention Trial. Findings related to a priori hypotheses of the trial. The Mul-tiple Risk Factor Intervention Trial Research Group. JAMA 1990;263(13):1795-801.

207. Persistence of reduction in blood pressure and mortality of participants in the Hypertension Detection and Follow-up Program. Hypertension Detection and Follow-up Program Cooperative Group. JAMA 1988;259(14):2113-22.

208. McGovern PG, Jacobs DR, Jr., Shahar E, Arnett DK, Folsom AR, Blackburn H, et al. Trends in acute coronary heart disease mortality, morbidity, and medi-cal care from 1985 through 1997: the Minnesota heart survey. Circulation2001;104(1):19-24.

209. Frank CW, Weinblatt E, Shapiro S. Angina pectoris in men. Prognostic signifi-cance of selected medical factors. Circulation 1973;47(3):509-17.

210. Kannel WB, Feinleib M. Natural history of angina pectoris in the Framingham study. Prognosis and survival. Am J Cardiol 1972;29(2):154-63.

211. Sigurdsson E, Sigfusson N, Agnarsson U, Sigvaldason H, Thorgeirsson G. Long-term prognosis of different forms of coronary heart disease: the Reykjavik Study. Int J Epidemiol 1995;24(1):58-68.

212. Rosengren A, Wilhelmsen L, Hagman M, Wedel H. Natural history of myocar-dial infarction and angina pectoris in a general population sample of middle-aged men: a 16-year follow-up of the Primary Prevention Study, Goteborg, Sweden. J Intern Med 1998;244(6):495-505.

213. Lampe FC, Whincup PH, Wannamethee SG, Shaper AG, Walker M, Ebrahim S. The natural history of prevalent ischaemic heart disease in middle-aged men. Eur Heart J 2000;21(13):1052-62.

214. Zethelius B, Byberg L, Hales CN, Lithell H, Berne C. Proinsulin and acute insulin response independently predict Type 2 diabetes mellitus in men--report from 27 years of follow-up study. Diabetologia 2003;46(1):20-6. Epub 2003 Jan 8.

215. Walldius G, Jungner I. Apolipoprotein B and apolipoprotein A-I: risk indicators of coronary heart disease and targets for lipid-modifying therapy. J Intern Med2004;255(2):188-205.

216. Marcovina SM, Albers JJ, Kennedy H, Mei JV, Henderson LO, Hannon WH. International Federation of Clinical Chemistry standardization project for meas-urements of apolipoproteins A-I and B. IV. Comparability of apolipoprotein B values by use of International Reference Material. Clin Chem 1994;40(4):586-92.

73

217. Marcovina SM, Albers JJ, Henderson LO, Hannon WH. International Federa-tion of Clinical Chemistry standardization project for measurements of apolipo-proteins A-I and B. III. Comparability of apolipoprotein A-I values by use of in-ternational reference material. Clin Chem 1993;39(5):773-81.

218. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne-Parikka P, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med2001;344(18):1343-50.

219. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Reduction in the incidence of type 2 diabetes with lifestyle interven-tion or metformin. N Engl J Med 2002;346(6):393-403.

220. Haffner SM, Greenberg AS, Weston WM, Chen H, Williams K, Freed MI. Effect of rosiglitazone treatment on nontraditional markers of cardiovascular disease in patients with type 2 diabetes mellitus. Circulation 2002;106(6):679-84.

221. Beckman J, Raji A, Plutzky J. Peroxisome proliferator activated receptor gamma and its activation in the treatment of insulin resistance and atherosclero-sis: issues and opportunities. Curr Opin Cardiol 2003;18(6):479-85.

222. Wald NJ, Law MR. A strategy to reduce cardiovascular disease by more than 80%. BMJ 2003;326(7404):1419.

Acta Universitatis UpsaliensisComprehensive Summaries of Uppsala Dissertations

from the Faculty of MedicineEditor: The Dean of the Faculty of Medicine

Distribution:Uppsala University Library

Box 510, SE-751 20 Uppsala, Swedenwww.uu.se, [email protected]

ISSN 0282-7476ISBN 91-554-5980-3

A doctoral dissertation from the Faculty of Medicine, Uppsala University,is usually a summary of a number of papers. A few copies of the completedissertation are kept at major Swedish research libraries, while the sum-mary alone is distributed internationally through the series Comprehen-sive Summaries of Uppsala Dissertations from the Faculty of Medicine.(Prior to October, 1985, the series was published under the title “Abstracts ofUppsala Dissertations from the Faculty of Medicine”.)

clinical manifestations of coronary heart disease and - diva portal

Documents