Oral glucose tolerance test as a prognostic tool in patients with acute coronary syndrome

(1)

Karolinska Institutet, Department of Medicine Cardiology Unit, Karolinska University Hospital

Stockholm, Sweden

ORAL GLUCOSE TOLERANCE TEST AS A PROGNOSTIC TOOL IN PATIENTS WITH ACUTE CORONARY SYNDROME

Catarina Djupsjö

M.D.

Stockholm 2022

(2)

Ca

All previously published papers were reproduced with permission from the publisher.

Published by Karolinska Institutet.

Printed by Universitetsservice US-AB, 2022

Cover illustration: Hjärtan-Gelé, Pixabay License

(3)

Oral glucose tolerance test as a prognostic tool in patients with acute coronary syndrome

By Catarina Djupsjö

THESIS FOR DOCTORAL DEGREE (Ph.D.)

The thesis will be defended in public at Torsten Ghord (S2:02) at Norrbacka, Karolinska University Hospital, Stockholm, Friday May 13^th at 9 am.

Principal Supervisor:

Jeanette Kuhl, MD, Ph.D., Associate Professor Karolinska Institutet

Department of Medicine, Solna Division of Cardiology

Danderyd University Hospital Division of Medicine

Co-supervisor(s):

Thomas Nyström, MD, Professor, Karolinska Institutet

Department of Clinical Science and Education Södersjukhuset

Division of Endocrinology

Magnus Lundbäck, MD, PhD., Associate Professor

Karolinska Institutet, Danderyd Department of Clinical Sciences Danderyd University Hospital Division of Cardiology

Opponent:

Annica Ravn-Fischer, MD, PhD., Associate Professor

Göteborg’s University

Department of Sahlgrenska Akademin Sahlgrenska University Hospital Division of Cardiology

Examination Board:

Claes Held, MD, Professor, Uppsala University

Department of Medical Science Division of Cardiology

Stefan Söderberg, MD, Professor, Umeå University

Department of Public Health and Clinical Medicine Division of Medicine

Buster Mannheimer, MD, PhD., Associate professor

Karolinska Institutet

Department of Clinical Science and Education Södersjukhuset

Division of Internal Medicine

(4)

(5)

“The only true wisdom is in knowing you know nothing.”

Socrates

(6)

(7)

ABSTRACT

Background: Disturbances of glucose metabolism such as type 2 diabetes and impaired glucose tolerance are established risk factors for cardiovascular disease and mortality. The disturbances lead to hyperglycaemia and the most common methods to diagnose hyperglycaemia are HbA1c, oral glucose tolerance test (OGTT) and fasting plasma glucose (FPG). The best method to predict death and cardiovascular disease is still being debated.

Aims:

1. To assess cardiovascular outcome a decade after acute coronary syndrome (ACS) and its relationship with repeated measurements of metabolic status and indices of glycaemic abnormalities.

2. To study long-term prognosis in patients with acute myocardial infarction and glucose abnormalities and focus on the predictive value of an oral glucose tolerance test and HbA1c.

3. To study preoperative disturbances of glucose metabolism and mortality after coronary artery bypass grafting (CABG).

4. To study the prognostic importance of random prandial blood glucose tests and the risk of mortality and cardiovascular events in patients attending the Emergency Department (ED).

Methods: In studies I-III all participants without known diabetes underwent a 75-g OGTT according to the World Health Organization (WHO) and were divided into different glucose tolerance groups depending on the result of the OGTT. In studies I-II, the OGTT was performed 4-5 days after the study population had been treated for myocardial infarction. In paper III, the OGTT was performed within 3 months of a CABG. In study IV, the study population had a random blood glucose test taken upon admission to the ED.

Results: In Study I, the revaluation of the metabolic status in 515 patients was done after 4 years and the mortality and cardiovascular events were studied after a decade. Patients with known diabetes type 1 and type 2 had higher mortality and incidence of cardiovascular events compared to the patients without diabetes, and patients with prediabetes had a higher incidence of cardiovascular events compared to patients with diabetes diagnosed by OGTT. At 4-years follow-up, the indices of metabolic control were higher in patients with dysglycaemia than in patients with normal glucose tolerance (NGT).

In Study II, 754 patients out of 1684 treated for acute myocardial infarction had an HbA1c and an OGTT controlled at the time of their infarction and were diagnosed with dysglycaemia using either OGTT or HbA1c. HbA1c in the prediabetes range, but not OGTT, added predictive value on the long-term outcome.

In Study III, 497 patients were studied regarding their outcome in accordance with their glucose tolerance group after CABG. There was no significant difference in all-cause mortality between patients with diabetes, prediabetes or NGT during a mean follow-up time of 10 years.

In Study IV, 662,018 patients had a random plasma glucose test when attending the ED.

Patients with newly discovered diabetes had approximately a 2-fold risk of all-cause mortality compared to patients with normal glucose tolerance at 3.9 years follow-up.

Conclusion: A majority of patients with cardiovascular events have undiagnosed disturbances of glucose tolerance and the long-term outcome in those patients is poorer compared to patients with NGT. HbA1c may be a better predictor than OGTT to identify patients at risk for premature death and cardiovascular events. Elevated plasma glucose test taken in the ED, can predict premature death.

(8)

(9)

LIST OF SCIENTIFIC PAPERS

I

Majid Kalani, Catarina Djupsjö, Gun Jörneskog, Jeanette Kuhl

Poor cardiovascular outcome a decade after acute coronary syndrome in patients with impaired glucose metabolism

Submitted

II

Stelios Karayiannides/ Catarina Djupsjö, Jeanette Kuhl, Claes Hofman-Bang, Anna Norhammar, Martin J Holzmann, Pia Lundman

Long-term prognosis in patients with acute myocardial infarction and newly detected glucose abnormalities: predictive value of oral glucose tolerance test and HbA1c

Cardiovascular Diabetology, 20:122, 2021

III

Catarina Djupsjö, Ulrik Sartipy, Torbjörn Ivert, Stelios Karayiannides, Pia Lundman, Thomas Nyström, Martin J. Holzmann, Jeanette Kuhl

Preoperative disturbances of glucose metabolism and mortality after coronary artery bypass grafting

Open Heart 7:e001217, 2020

IV

Catarina Djupsjö, Jeanette Kuhl, Magnus Lundbäck, Martin J. Holzmann, Thomas Nyström Admission glucose as a prognostic marker for all-cause mortality and cardiovascular

disease Submitted

(10)

(11)

LIST OF ABBREVIATIONS

ACCORD Action to Control Cardiovascular Risk in Diabetes

ACS Acute coronary syndrome

ADA American Diabetes Association

ADVANCE In the Action in Diabetes and Vascular Disease-PreterAx and DiamicroN Controlled Evaluation

AMI Acute Myocardial Infarction

BMI Body mass index

CABG Coronary artery bypass grafting

CAD Coronary artery disease

CVD Cardiovascular disease

CI Confidence Intervals

DECODE The Diabetes Epidemiology: Collaborative analysis of Diagnostic criteria in Europe

EASD European Association for the Study of Diabetes

ED Emergency Department

ESC European Society of Cardiology

FFA Free Fatty Acid

FPG Fasting Plasma Glucose

GLP-1 glucagon-like peptide-1

HbA1c Glycated haemoglobin A1c

IFCC International Federation of Clinical Chemistry Working Group

IFG Impaired fasting glucose

IGT Impaired glucose tolerance

IR Insulin Resistance

JDS Japanese Diabetes Society

HR Hazard Ratio

LADA Latent Autoimmune Diabetes in Adults

LDL Low-Density Lipoproteins cholesterol

MI Myocardial Infarction

MODY Maturity-Onset Diabetes of the Young

NGSP National Glycohemoglobin Standardization Program

NGT Normal glucose tolerance

NPR National Patient Register

NSTEMI Non-ST-elevation Myocardial Infarction

OGTT Oral glucose tolerance test

PCI Percutaneous coronary intervention

SCAAR Swedish Coronary Angiography and Angioplasty Registry

SD Standard deviation

SEPHIA National Registry of Secondary Prevention SGLT2 Sodium-glucose co-transporter

STEMI ST-Elevation Myocardial Infarction

SWEDEHEART Swedish Web System for Enhancement and Development of Evidence-based Care in Heart Disease Evaluated According to Recommended Therapies Registry

TIA Transient Ischemic Attack

T1DM Type 1 diabetes mellitus

T2DM Type 2 diabetes mellitus

UKPDS The UK Prospective Diabetes Study VADT the Veterans Affairs Diabetes Trial

WHO World Health Organization

(14)

(15)

1 INTRODUCTION

1.1 BACKGROUND

Atherosclerosis is the main reason for cardiovascular disease (CVD) and it is chronic inflammation where fatty substances form plaque around the artery walls, reducing both the blood flow and the oxygen supply to the myocardium (1). Diabetes mellitus is an important risk factor in the development of atherosclerosis and CVD (2, 3). Patients with known diabetes have a higher incidence of acute myocardial infarction (AMI) with a higher rate of long-term mortality compared to patients with normal glucose tolerance (NGT) (4). Less is known about the incidence of myocardial infarction (MI) and long-term mortality in patients with early stages of disturbances of glucose metabolism, also known as prediabetes. Approximately two- thirds of patients with MI have unknown and untreated disturbances of glucose metabolism (5). There are conflicting data regarding the impact of prediabetic status and the long-term cardiovascular prognosis following MI. Some data show that patients with prediabetes have the same prognosis as patients with normal glucose tolerance while other studies can show that patients with prediabetes have a worse outcome (6-8).

It is important to identify the long-term prognosis in patients with CVD and disturbances of glucose metabolism and which method is best used for this purpose. Type 2 diabetes (T2DM) and prediabetes can mainly be diagnosed by different methods: fasting plasma glucose (FPG), glycated haemoglobin A1c (HbA1c), random plasma glucose test and an oral glucose tolerance test (OGTT) (9). There has been a debate regarding the best method to diagnose disturbances of glucose metabolism following MI (4, 6-8, 10-12). The European Society of Cardiology (ESC) stated in their latest guidelines “Diabetes, prediabetes and cardiovascular disease” that FPG and/or HbA1c should be used first and OGTT only if these tests are inclusive. The guidelines also stated that there is a gap in evidence in the predictive abilities of HbA1c vs OGTT for hard outcomes in patients with cardiovascular disease (13).

1.2 CARDIOVASCULAR DISEASE 1.2.1 General

CVD usually refers to coronary artery disease (CAD), peripheral artery disease (PAD) and cerebrovascular disease (14).

CAD consists of acute coronary syndrome (ACS), stable angina and/or heart failure due to ischemic damage. ACS is a state of reduced myocardial perfusion which causes MI and unstable angina pectoris (15). The clinical presentation of MI and unstable angina pectoris is often similar to chest pain which can radiate out to the left arm, the back and to the jaw. Other symptoms are dyspnoea, nausea, sweating or fatigue. The symptoms are a result of a reduced blood flow in the coronary arteries, usually due to atherosclerotic plaques or thromboses that block or narrow the arteries. An MI, is often divided into non-ST-elevated myocardial infarction (NSTEMI) and ST-elevated myocardial infarction (STEMI), the ischaemia is so severe it causes myocardial damage, often myocardial necrosis and the release of biomarkers such as Troponin T and Troponin I (15). The loss of viable myocardium affects cardiac output and can result in heart failure, formation of a left ventricular aneurysm, rhythm disturbances as well as ventricular septal rupture and papillary muscle rupture (16). Unstable angina pectoris

(16)

is considered a pre-stage of MI and is an acute condition for the first 3-4 weeks it appears since it is uncertain if it will develop into MI. Angina pectoris that does not progress and lasts for more than 4 weeks is considered a more chronic form and is then referred to as stable angina pectoris (15). CAD is also the most common cause of heart failure with a reduced ejection fraction (17). The initial ischaemic damage to the myocardium by the reduced perfusion of blood results in necrosis of the myocytes and remodelling. The most common symptoms are dyspnoea, orthopnea and oedema, tachycardia and pleura effusion (17).

PAD affects the lower extremities due to atherosclerosis which narrows the arteries. Typical symptoms are claudicatio intermittens, numbness of the limbs, rest pain, critical ischaemia with ulcers in the lower extremities which by time can develop into gangrene if untreated (18).

CVD affects the arteries and the blood circulation of the brain, and the most common presentation is stroke and transient ischaemic attack (TIA). Stroke symptoms usually have an acute onset, with different neurological symptoms such as hemiplegia, aphasia, ataxia or affected vision depending on where in the brain the stroke is located. A TIA resembles a stroke in symptoms and onset but the reduced perfusion or blockage of the arteries is quickly restored leaving no brain damage or remaining symptoms within 24 h (19). Approximately 85% of strokes are ischaemic and the rest are haemorrhagic (20). Hypertension is the most important risk factor in developing a stroke. Atrial fibrillation/flutter, diabetes, dyslipidaemia and smoking are other important risk factors (19).

1.2.2 Epidemiology of cardiovascular disease

CVD causes approximately 17.9 million deaths each year and is the number one cause of death globally (21). In Europe, CVD causes 3.9 million deaths each year and accounts for 45% of all deaths in Europe. CVD is the main cause of death in men in all but 12 countries and the main cause of death in women in all but two countries.

In Sweden, the prevalence of MI has decreased during the last 20 years. In 2018, 24,800 people were treated for MI compared to 36,400 in 1998 (22). The reduction is mainly due to a combination of advanced medical treatment of risk factors such as hypertension, elevated cholesterol levels, improvement of lifestyle factors with a decrease in smoking habits and improved emergency medical care of patients with MI (23).

1.2.3 Pathophysiology of atherosclerosis

Atherosclerosis is the primary cause of CVD, and it affects the large, medium-sized and small arteries (1). The pathophysiology of atherosclerosis commences when Low-Density Lipoproteins (LDL) accumulate in the subendothelial wall, leading to chronic inflammation affecting the endothelial barrier by recruiting T-cells and monocytes to the site. It is possible that the oxidation of LDL cholesterol particles in the intima wall could be the signal to the inflammatory cells but the process is not known in detail (1). The monocytes differentiate into macrophages and develop receptors that can bind the oxidized LDL cholesterol particles creating “foam cells” leading to chronic inflammation. When the foam cells are overloaded, the cholesterol is accumulated and forms atherosclerotic plaques. These plaques can both narrow the coronary vessels preventing a sufficient supply of oxygenated blood to the heart and/or rupture causing a complete blockage, both situations may result in MI (1).

(17)

Late complications of atherosclerosis by Npatchett

https://commons.wikimedia.org/wiki/File:Late_complications_of_atherosclerosis.PNG https://creativecommons.org/licenses/by-sa/4.0, via Wikimedia Commons

1.2.4 Risk factors for cardiovascular disease with a focus on coronary artery disease

Diabetes, smoking, hypertension, lack of exercise, obesity, diet and elevated blood lipid levels are all risk factors for MI. The effect of potentially modifiable risk factors associated with MI in 52 countries (the INTERHEART study) demonstrated that these are the major risk factors for MI in both men and women regardless of age and nationality (24).

Diabetes is an established risk factor for CAD and patients with diabetes have substantially higher mortality in CAD than non-diabetic patients (25). An association between intracellular metabolic changes and hyperglycaemia is believed to enhance atherosclerosis due to oxidative stress, endothelial dysfunction and inflammation (26). The coexistence of other risk factors for diabetes i.e. obesity, hypertension, dyslipidaemia and chronic kidney disease might also contribute to an enhanced risk of CAD (27, 28).

Hypertension is defined as a systolic blood pressure value ≥ 140 mmHg and/or diastolic blood pressure value ≥ 90 mmHg in patients without diabetes and as systolic blood pressure value

≥ 130 mmHg and/or diastolic blood pressure value ≥ 80 mmHg in patients with diabetes and microalbuminuria. Patients with diabetes is often recommended, after treatment, to have a target blood pressure of ≤ 130 mmHg systolic pressure and ≤ 90 mmHg diastolic pressure (29).

An increase in systolic blood pressure will result in an increase of oxygen demand for the myocytes which can ultimately result in myocardial infarction. Hypertension also contributes to the development of atherosclerosis by increased stress in the coronary arteries, increased transmural pressure and wall stress (30).

Smoking is another risk factor for CAD. Cigarette smoke can increase the level of endothelial dysfunction and vascular damage and initiate processes that leads to atherosclerosis, such as the generation of cytokines, increased endothelial permeability, oxidative stress, impaired serum lipid profile and platelet aggregation (31). Smoking also affects platelet function, fibrinolytic factors and antithrombotic/pro-thrombotic factors leading to the formation of thrombosis and in a worst-case scenario, to ACS (32).

(18)

A lack of physical exercise increases the risk of developing CAD. Physical exercise refers to the frequency, intensity (percentage of heart rate or lung volume) and duration. Physical exercise temporarily increases insulin sensitivity and the insulin-independent glucose uptake in the muscles and can improve metabolic status (33). Laboratory testing and clinical findings have also shown that exercise influences blood coagulation, fibrinolysis, vascular remodelling, blood pressure and blood lipid profiles (34).

Obesity is defined by body mass index (BMI), a ratio between body weight to height (kg/m²).

A normal person has a BMI of 18.5-24.9 kg/m2 whereas an overweight person has a BMI of 25-29.9 kg/m² and an obese person has a BMI of > 30 kg/m². Obesity, particularly abdominal obesity, promotes the development of insulin resistance and increases the risk of hypertension, increased blood lipids and inflammation leading to atherosclerosis (34). The result of Insulin Resistance (IR), elevated triglycerides and hyperglycaemia can, through different methods such as inflammation, altered calcium signalling within the cells, mitochondrial dysfunction and the generation of toxic metabolic substances ultimately lead to hypertrophic cardiomyopathy and heart failure (35).

1.2.5 Definition and diagnosis of myocardial infarction and unstable angina pectoris

A combination of patient history, ECG, imaging and the cardiac biomarker troponin are used to diagnose MI and ACS. According to the Fourth Universal Definition of Myocardial Infarction by Thygesen et al (36), MI is defined as clinical symptoms of acute myocardial ischaemia, a rise and/or fall of biomarker troponin with at least one value above the 99^th percentile URL and one of the following: ischaemic ECG changes, development of pathological ECG Q waves, identification of a coronary thrombus by angiography or autopsy, imaging evidence of new regional wall motion abnormality or new loss of viable myocardium.

This is also referred to as MI type 1. Another common form of infarction is MI type 2, which is when oxygen supply and demand are not met leading to an ischaemic myocardial injury without underlying atherothrombosis, such as hypotension, anaemia, bleeding and shock (36).

Unstable angina pectoris has a similar definition as MI but does not cause cell damage and will therefore not have biomarkers released in the bloodstream (36).

1.3 DISTURBANCES OF GLUCOSE METABOLISM 1.3.1 General

Diabetes mellitus is a condition where the β-cells of the pancreatic islets or islets of Langerhans are unable to produce enough insulin to metabolize the body’s glucose intake, resulting in elevated blood glucose levels. Typical symptoms of diabetes are polyuria, increased thirst and hunger, weight loss and fatigue. Untreated elevated blood glucose levels (hyperglycaemia) over a long period of time causes both micro and macrovascular complications (37).

There are several forms of diabetes where diabetes type 1 (T1DM) and diabetes type 2 (T2DM) are the two most common forms. Patients with T1DM have no insulin production at all whereas patients with T2DM are resistant to insulin or have insufficient insulin production (38). Another form of diabetes is monogenic diabetes, which is caused by a defect of a single gene. The most common forms of monogenic diabetes are maturity-onset diabetes of the

(19)

young (MODY), syndromic diabetes and neonatal diabetes. The symptoms are similar to T1DM and T2DM but since it is caused by a specific gene the form is diagnosed by genetic testing and a more precise treatment can be achieved. Monogenetic diabetes accounts for approximately 1-5% of all diabetes (39). Diabetes can also be caused by a specific condition that destroys the β-cells and/or induces insulin resistance, such as endocrinopathies (e.g.

thyroid disease, Cushing syndrome, acromegaly) (40).

Diabetes is a broad definition, and the classification is heterogeneous. By looking into different risk factors one can divide diabetes into several subgroups with different patient characteristics and different diabetes complications which can help individualize treatment (41). Instead of categorizing T1DM and T2DM, Ahlqvist et al divided diabetes into six variables (BMI, age at diabetes diagnosis, HbA1c, glutamate decarboxylase antibodies, homeostatic model assessment 2 estimate β-cell function and insulin resistance (IR)) which identified five different clusters of patients which different patient characteristics and risk of diabetes complications (41).

Elevated glucose levels can also be due to stress which may not result in a diabetes diagnosis (42). This stress hyperglycaemia is explained by different physiological mechanisms than prediabetes and T2DM (43) such as IR caused by multiple neuroendocrine responses and can be a result of an acute illness. Stress hyperglycaemia has been suggested as an essential survival response (44) and is common in critically ill patients. It has been suggested that stress hyperglycaemia is a marker of disease severity (44).

1.3.2 Epidemiology of diabetes

The prevalence of diabetes is increasing in the world. In 2021, approximately 537 million people had diabetes and it is estimated that this figure will rise to 643 million in 2030 and 738 million in 2045 (45).

In 2016, WHO (World Health Organization) estimated that diabetes was the seventh leading cause of death in the world and the prevalence is rising more rapidly in middle and low- income countries (46). In Sweden, the prevalence of diabetes was 4.4% in 2013. However, it is estimated that it will rise to 10.4% by 2050 (47).

1.3.3 Pathophysiology of diabetes Diabetes type 1 (TIDM)

T1DM is often diagnosed at an early age and is also called juvenile diabetes and accounts for 5-10% of those with diabetes (48). T1DM can also occur later in life and is then referred to as latent autoimmune diabetes in adults (LADA). Patients with LADA are > 35 years, have positive antibodies to islet beta cells usually found in patients with T1DM, e.g. Glutamic acid decarboxylase antibodies (GADA antibodies) and a low reference value of C-peptide level. C- peptide is released from the β-cells and is a measurement of the body’s insulin production (49).

Approximately 40% of T1DM occurs in patients older than 30 years (50). It is caused through autoimmune destruction of β-cells in the pancreas leading to an absolute insulin deficiency.

The autoimmune disorder and other risk factors for developing the disease are believed to be the presence of certain genes, a family history, geography/ethnicity, infections and age. But the reason for the destruction of the beta cells is not fully clarified (51).

(20)

Diabetes type 2 (T2DM)

T2DM, also known as adult-onset diabetes and accounts for 90-95% of diabetic patients and is especially prevalent in the elderly and the obese (48, 52) The most common risk factors for developing T2DM are obesity, lack of exercise, smoking, a family history, gestational diabetes and hypertension where obesity and lack of physical activity are the most prominent factors (1, 53). With food intake, normally more energy is consumed than needed and some of that energy, i.e. fat and carbohydrates, is used immediately by the metabolism as fuel and the excess is stored in either adipocytes or as glycogen in the muscle and hepatic cells for later use. For the transportation of glucose to these cells and fatty acid uptake, insulin is mandatory (54). In T2DM hyperglycaemia results from a combination of insulin resistance and impaired insulin secretion (38).

- Insulin resistance (IR)

IR is the most important factor in T2DM and is a condition where normal insulin levels are not enough to prevent elevated blood glucose and is most noticeable in skeletal muscle, in the liver and as a dysfunction of the β-cells (55).

The cells of the skeletal muscles have the largest glucose uptake (normally 80% of a carbohydrate load) and in IR there is reduced glucose uptake in the muscles following a meal which causes hyperglycaemia. IR leads to increased lipolysis in the adipose tissue resulting in elevated levels of free fatty acids in the blood which in turn reduces the glucose uptake and the glucose metabolism in the skeletal muscles (54). In the liver, glucose is produced during periods of fasting when the blood glucose levels are low. This production is prevented by hyperglycaemia and commences by hypoglycaemia and elevated glucagon levels (56). Due to IR, glucose production in the liver is not halted by elevated glucose levels, but continues, and this is resulting in hyperglycaemia both during periods of fasting as well as during meals (55).

The β-cells are normally able to increase their insulin secretion as a response to hyperglycaemia which will result in normal glucose levels. If hyperglycaemia is consistent, β-cells will be exhausted and can begin to falter which will result in elevated glucose levels and diabetes. IR is often referred to as the triumvirate, which is a combination of the IR mechanisms in skeletal muscles and hepatic cells, accelerated lipolysis in adipose cells and β-cell failure (55).

IR begins several years before the diagnosis of diabetes (57). It is caused by both genetic and environmental factors resulting in the reduced effect of insulin where obesity is one of the most dominant causes (58, 59). Genetic factors include abnormalities of the enzymes involved with insulin activity within the cells or of the insulin receptor on the surface of the cells which means that insulin cannot activate the necessary reactions within the cells. In particular, central obesity has been linked to IR as a result of increased oxidative stress, higher production of pro- inflammatory adipocytokines and activation of the renin-angiotensin-aldosterone system (59, 60).

- Impaired insulin secretion

The insulin secretion from the β-cells is regulated through blood glucose concentration, and hyperglycaemia triggers insulin secretion. In patients with T2DM, insulin secretion is lower than in patients without diabetes and can be a result of several different factors. Impaired insulin secretion can be a result of IR leading to hyperinsulinemia of the β-cells and with time,

(21)

disorders of the β-cells, such as β-cell secretion, differentiation and the growth of the β-cells which can also lead to a decrease in insulin secretion (54).

β-cell function can also be disturbed by elevated levels of glucose and triglycerides, commonly referred to as glucotoxicity and lipotoxicity (61). Elevated glucose levels are toxic for the β- cells and can cause damage and loss of function which ultimately leads to a decrease in insulin production and a reduced insulin secretion. This leads to even higher glucose levels and a vicious cycle is started. A similar cycle is caused by triglycerides (54, 61). Increased glucotoxicity and lipotoxicity will result in the loss of β-cell function and it is estimated that approximately 50% of β-cell insulin production is lost when T2DM is diagnosed (62).

A schematic diagram illustrating insulin, glucose, and Free Fatty Acid (FFA) regulation after a meal by Anarina L. Murillo et al

https://images.app.goo.gl/Quokks3PvABXErWf8

Copyright: 2019 A.L. Murillo, J. Li, C. Castillo-Chavez, licensee AIMS Press

1.3.4 Complications of diabetes and prediabetes

Diabetes and prediabetes affect the body in many different ways and have a direct effect on both microvascular and macrovascular levels and therefore a great impact on mortality and morbidity. Macrovascular complications refer to the larger blood vessels of the brain, heart and legs and are mainly due to atherosclerosis. Microvascular complications refer to the smaller vessels, and microvascular complications refer to the capillaries all over the body where the kidneys and eyes are the most affected organs. Microvascular complications of diabetes may start early and can occur up to seven years before the diagnosis of T2DM (63).

(22)

The most common forms of microvascular complications are:

• Retinopathy – The most common form of microvascular complication. Retinopathy is the damage of the small vessels in the retina which leads to blurred vision or blindness if untreated.

• Nephropathy – A chronic loss of renal function due to damage to glomeruli in the kidney which leads to glomerular hyperfiltration and albuminuria due to renal damage.

• Neuropathy – Damage to the peripheral nerves leading to numbness, pain, tingling or weakness in the extremities.

The most common forms of macrovascular complications are:

• CVD – Narrowing of the coronary arteries due to atherosclerosis and the formation of plaques/thrombosis resulting in angina or MI. The result of MI can lead to damage of the myocytes and heart failure with a reduced ejection fraction.

• Cerebrovascular disease – Narrowing of the arteries in the cerebellum or the carotid arteries due to atherosclerosis and the formation of plaques and/or thrombosis which can result in a transient ischaemic attack or stroke.

• PAD – Narrowing of the peripheral arteries due to atherosclerosis in the lower extremities reducing the oxygenated blood flow to the extremities resulting in intermittent claudication and/or skin ulcers. The worst complication is tissue damage which may require amputation.

Microvascular and macrovascular complications can occur by the presence, or even before the diagnosis of diabetes, and by risk factors such as hypertension, obesity, dyslipidaemia, albuminuria or smoking (64). Observational studies show that patients with T2DM and no other risk factors, as above, have a similar risk of cardiovascular mortality and MI as patients without disturbances of glucose metabolism if well treated (64).

Diabetes and cardiovascular disease

A cardiovascular event is an important complication of diabetes and the most common cause of premature death in patients with diabetes (65). Diabetes mellitus is an established risk factor for CVD and mortality and patients with diabetes have a worse outcome following an acute coronary event (66). The excess risk of mortality and a cardiovascular event increases with the level of disturbances of glucose metabolism (67). In contrast, recently it has also been shown that even prediabetes is also a risk factor for cardiovascular events and mortality which can occur even before the onset of diabetes (68-70).

Furthermore, cardiovascular complications can appear at the same time as diabetes is diagnosed. It is common to find previously undiagnosed prediabetes and diabetes in patients

(23)

admitted for an acute MI (71). Approximately 2/3 of patients admitted with an MI have previously unknown disturbances of glucose metabolism where hyperglycaemia has shown to be persistent at follow-up after 3 months (72, 73). A similar situation can be found in stroke patients where approximately 2/3 of the patients had unknown hyperglycaemia when admitted for stroke but this hyperglycaemia was less noticeable at follow-up after three months (74).

Patients with diabetes also have a higher risk of developing heart failure and the prognosis for patients with diabetes and heart failure is poorer compared to patients without diabetes (75, 76). One large study could show that approximately 44% of patients hospitalized due to heart failure have diabetes (77).

In patients with diabetes and multivessel CAD, the recommended revascularization treatment is CABG (78-81). CABG in patients with T2DM and multivessel disease still have a better mortality rate when compared to PCI but the prognostic outcome in patients with known T2DM is poorer compared to patients with normal glucose tolerance after CABG (82-84). A few studies have managed to show similar long-term mortality after CABG in patients with known T2DM and patients without diabetes (6). Less is known about the prognostic importance of prediabetes or newly discovered diabetes in patients with no previous history of disturbances of glucose metabolism after a CABG operation. The relationship between prediabetes and newly discovered diabetes and the prognosis after CABG is still unclear. It was recently shown that patients with prediabetes or newly discovered diabetes prior to CABG had similar long-term mortality compared to patients with normal glucose tolerance, but further studies are needed (85).

Treatment of dysglycaemia

To prevent complications of diabetes, treatment of hyperglycaemia is important. Both the American Diabetes Association (ADA) and European Association for the Study of Diabetes (EASD) have recommended guidelines for the treatment of diabetes where lifestyle modifications, pharmaceutical treatments and treatment of CVD risk factors play an important role to reach the recommended target glucose levels (86). Still, patients with T2DM are at higher risk of CVD compared to patients without diabetes.

Several studies have investigated whether more intensive treatment of patients with T2DM and the risk factors for cardiovascular disease can reduce the risk of cardiovascular events and mortality. The UK Prospective Diabetes Study (UKPDS) study showed that intensive blood glucose treatment with insulin or sulphonylureas had no risk-reduction of macrovascular complications but reduced the risk of microvascular complications in comparison to the group with conventional treatment (83). The more intensive treatment group also resulted in more episodes of hypoglycaemia compared to the conventional group (87). The Veterans Affairs Diabetes Trial (VADT) study investigated whether more intensive glycaemic control in patients with T2DM and risk factors for T2DM had a positive effect in comparison to a standard treatment group. Similar medications were used in both groups but in the intensive treatment group, the doses were increased since lower HbA1c was aimed for. The outcome shown no significant difference in CVD events and mortality between groups (88). The Action to Control Cardiovascular Risk in Diabetes (ACCORD) study evaluated whether a more intense treatment of dysglycaemia in combination with a more intense treatment of

(24)

hypertension or dyslipidaemia could reduce the risk of CVD events in patients with T2DM and risk factors for CVD in a 2x2 factorial design study. The findings of the study were that 2 different lipid therapies or more intensive treatment for hypertension in combination with an intensive glucose treatment did not reduce CVD in patients with T2DM. The intensive blood glucose-lowering therapy arm revealed an increased risk of mortality compared to the less intensive strategy and the study was stopped prematurely (89). In the Action in Diabetes and Vascular Disease-PreterAx and DiamicroN Controlled Evaluation (ADVANCE) study the vascular outcomes in T2DM were studied by a more intensive glucose treatment which resulted in a reduction of major macrovascular and microvascular events as a consequence of a reduction in nephropathy. There was no significant reduction in macrovascular events, and hypoglycaemia was present in the intensive treatment group (90). One smaller epidemiological study could show that patients with MI and hyperglycaemia treated with metformin had an improved outcome in comparison to patients treated with insulin, and further randomized controlled studies are recommended (91).

In recent years, new antidiabetic treatments have appeared that have produced a significant reduction of CVD in patients with T2DM. Sodium-glucose co-transporter (SGLT2) inhibitors was a new antidiabetic treatment where it recently has been shown that it can reduce the risk of cardiovascular events, mortality and renal outcomes in patients with T2DM and this is now a recommended treatment (92, 93). Also, glucagon-like peptide-1 (GLP-1) receptor agonists affect cardiovascular outcomes. Treatment with GLP-1 receptor agonists in patients with T2DM reduces both cardiovascular mortality and renal outcomes (94).

Another challenge is to screen patients with CVD for hyperglycaemia and the management of risk factors. Regardless of guideline recommendations, the screening for disturbances of glucose metabolism in patients with CVD is poorly managed. Even the management of pharmaceutical treatment and lifestyle modifications according to guidelines have not proven to be successful (95).

1.3.5 Diagnostic methods to detect disturbances of glucose metabolism In the old days, diabetes was diagnosed by sniffing or tasting the urine for a sweet smell or taste. Today, there are more modern versions where the following methods are approved methods by the WHO (96) and ADA (38).

• Fasting plasma glucose (FPG)

• Oral glucose tolerance test (OGTT)

• Haemoglobin A1c (HbA1c)

• Random prandial blood glucose testing and symptoms of dysglycaemia

For a long time, FPG and OGTT were the only two recommended methods to diagnose disturbances of glucose metabolism according to the WHO and ADA. In 2010, HbA1c was accepted by ADA (37) and in the following year by the WHO (9). ADA recommends all three tests at the same level. The WHO and ESC recommend primarily FPG or HbA1c and OGTT if a diagnosis is still in doubt (13).

(25)

Fasting Plasma Glucose (FPG)

A fasting plasma glucose measures a person’s glucose levels after fasting for a minimum of 8 hours. Two elevated measurements are needed to diagnose diabetes (38).

What is considered a normal FPG has been studied but with different results. Using a clinical approach, it is measured where the risk of mortality and cardiovascular disease start which is at an FPG of 4.5-6.09 mmol/L (97). Since there are inadequate details as to how to define a normal FPG, the WHO recommends that normal glucose levels should be correlated with a low risk of CVD or developing diabetes (96).

FPG was for a long time the favoured method to diagnose disturbances of glucose metabolism due to its convenience, affordable cost and availability in most laboratories throughout the world. The Diabetes Epidemiology: Collaborative Analysis of Diagnostic Criteria in Europe (DECODE) study and the Hoorn study could show that patients with diabetes diagnosed with FGP had a better outcome compared to patients diagnosed by OGTT (98, 99). But FPG alone identifies all patients with disturbances to their glucose metabolism and can only identify approximately 70% of people with diabetes (98, 100).

Oral Glucose Tolerance Test (OGTT)

OGTT is a method to identify diabetes, insulin resistance and impaired β-cell function. The patient is fasting before intake of a standardised glucose solution of 75 g anhydrous glucose dissolved in water. A fasting plasma glucose is measured before the intake of the glucose solution and another plasma glucose is measured after 2 hours have passed.

OGTT is a more time-consuming, expensive and less reproducible method than FGP and can diagnose 30% more patients with diabetes than patients diagnosed by FPG (98, 100). Patients diagnosed with disturbance of glucose metabolism by an OGTT have higher all-cause mortality than patients identified by FPG alone (98, 99). Impaired glucose tolerance, i.e.

normal FPG but an elevated 2-hour post-load glucose, can only be diagnosed by OGTT.

HbA1c

Haemoglobin, Hb, is an oxygen-transporting protein in the erythrocytes. Humans have a Hb form called Haemoglobin A which consists of two α- and two β- chains (101). In case of excessive glucose in the bloodstream, a monosaccharide is linked, glycated, to the β- chains of the Hb and creating HbA1c. HbA1c reflects the average levels of glucose in the bloodstream over the last 2-3 months (101). HbA1c is affected by conditions with a high turnover of erythrocytes such as anaemia, blood transfusion chronic liver disease, by altered haemoglobin such as haemoglobinopathies, by erythrocyte destruction such as splenectomy and hemoglobinopathies (101).

HbA1c can be measured any time of the day, is not sensitive to day-to-day variability and does not involve fasting. It is as specific and sensitive as FPG and OGTT in the diagnosis of diabetes (102) but requires a test method that is not available in all countries; in addition, the consistency of measurement differs across the world. The International Federation of Clinical Chemistry Working Group (IFCC) has developed reference methods for the analysis of HbA1c through the use of different designated comparison methods, such as the National Glycohemoglobin Standardization Program (NGSP), Japanese Diabetes Society (JDS) and the Mono-S method used in Sweden. IFCC is now the international reference method and is accepted by the WHO (103).

(26)

In comparison to OGTT and FPG, HbA1c is a more convenient method where fasting before blood sampling is not required. HbA1c is also associated with fewer day-to-day disturbances due to acute stress and has greater preanalytical stability (104).

Random prandial plasma glucose

A random prandial plasma glucose test can be used to diagnose diabetes. An elevated measurement and symptoms of diabetes are needed for the diabetes diagnosis (38). The test is not affected by a recent meal and does not require fasting. The method is better at diagnosing diabetes in people with a high blood glucose level than mildly elevated levels. It is also useful when an immediate or urgent diagnosis is needed, as in patients with newly discovered T1DM (105).

1.3.6 Classification of diabetes and prediabetes

Impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) are conditions with elevated blood glucose but not high enough to reach the criteria of diabetes. IFG is a result of impaired basal and glucose-stimulated first-phase insulin secretion and IGT is mainly due to peripheral insulin resistance and impaired second-phase insulin resistance (106).

Diabetes, IFG and IGT are classified according to WHO and ADA criteria (9, 37, 96):

Diabetes definition

• FPG ≥ 7.0 mmol/L is defined as T2DM by WHO and ADA

• Oral glucose tolerance test - FPG level ≥ 7.0 mmol/l, and/or a 2-h plasma glucose value at OGTT ≥11.1 mmol/l is defined as T2DM by WHO and ADA

• HbA1c – 48 mmol/mol or higher (6.5% DCCT) is defined as T2DM

• Random prandial plasma glucose value ≥11.1 mmol/l and symptoms of diabetes

Prediabetes definition:

• IFG is defined by impaired fasting glucose with a fasting plasma glucose value of ≥ 6.1 to < 7.0 mmol/l by WHO and ≥ 5.6 to < 7.0 mmol/L by ADA

• IGT is defined as impaired glucose tolerance with a fasting plasma glucose value of

<6.1 mmol/l by WHO and < 5.6 mmol/L by ADA and a 2-h plasma glucose level at OGTT ≥ 7.8 mmol/l to < 11.1 mmol/l by ADA and WHO

• HbA1c 39-47 mmol/mol (5.7-6.4% DCCT) by ADA, 42-47 mmol/mol (6.0-6.4%

DCCT) by WHO

• Random prandial plasma glucose level ≥ 7.8 to < 11.1 mmol/l

(27)

Definition of Impaired Fasting glucose by ADA and WHO

In defining IFG, both WHO and ADA had initially agreed on a glucose level cut-off point of 6.1 mmol/l as the upper limit due to the factors mentioned earlier, i.e. the increased loss of insulin sensitivity and ß-cell function and a higher risk of mortality and CVD (97, 99). In 2003, ADA changed the lower cut-off point for IFG to 5.6 mmol/l due to better specificity and sensitivity in predicting diabetes (107) even though the association with the risk of developing diabetes is greater with an FPG of 6.1-6.9 mmol/l than FPG 5.6-6-0 mmol/l (108).

The WHO remained with the lower cut-off point of 6.1 mmol/l for several reasons. One was that it had a better concordance with IGT where a greater number of patients with IFG also have IGT (109). It would also focus on the patients with the highest risk. Patients with IGT have a risk of higher all-cause and cardiovascular mortalityso a better concordance between IFG/IGT would help the health systems focus on patients most in need (99).

Definitions of normal glycaemic status and dysglycaemia according to ADA and WHO criteria

Glucose (mmol/L)

Random plasma glucose

and symptoms of diabetes HbA1c (mmol/mol)

FPG 2h-PG

Normal glycaemic status < 5.6 (ADA)

< 6.1 (WHO) < 7.8 < 39 (ADA)

< 42 (WHO) Impaired fasting glycaemia

(IFG)

5.6-6.9 (ADA)

6.1-6.9 (WHO) < 7.8

Impaired glucose tolerance (IGT)

<5.6 (ADA)

<6.1 (WHO) 7.8-11.0

Prediabetes 39-47 (ADA)

42-47 (WHO)

Diabetes mellitus type 2 ≥ 7.0 ≥ 11.1 ≥ 11.1 ≥ 48

Studies have shown that OGTT can identify more patients with diabetes and impairment of glucose metabolism than HbA1c and fasting plasma glucose respectively (110, 111).

However, it has not been clarified if the screening by OGTT in patients with ACS will affect their outcome regarding mortality and longevity or if it will be sufficient to screen these patients with Hb1Ac. The ESC is aware of the lack of consensus and information in this area and stated in its latest guidelines in diabetes, prediabetes and cardiovascular disease that

“Direct comparison of the predictive abilities of HbA1c- vs OGTT-derived measures for hard outcomes in people with CVD as gaps in evidence (13).

1.3.7 Random blood glucose as a prognostic marker for mortality

Stress-induced hyperglycaemia, which occurs in extreme situations, has another physiological explanation for elevated blood glucose levels than diabetes and prediabetes (43). Hormones, e.g. glucagon, glucocorticoids and catecholamine, the central nervous system and cytokines which are all active in stress-mediated situations interfere with the metabolism in liver and skeletal muscle resulting in impaired glucose uptake in the skeletal muscle and reduced suppression of hepatic gluconeogenesis (43).

Stress hyperglycaemia, measured as random blood glucose, in correlation to a specific condition, such as MI, stroke, heart failure or pneumonia, is associated with higher in-hospital

(28)

mortality, increased length of hospital stay and a higher rate of in-hospital complications (112- 120). The prognostic factor of random blood glucose in a general, in an acute population, in a population in a hospital, or a population not associated with a specific condition, is also associated with an increased length of hospital stay and mortality (121). The long-term effects of the prognostic factor of random glucose in an acute population are less studied. It was recently shown that patients with elevated glucose at an acute general medical ward, no matter the cause of admission, had higher mortality after one year, but not after two years (122).

1.4 REGISTERS

Epidemiological research in Sweden is possible due to many population-based national registries created for research purposes (123). Thanks to the unique personal identity numbers all citizens in Sweden have, data can be linked to a specific individual from different registries (124). In this thesis, the following registries have been used for data collection:

1.4.1 The National Patient Register

The National Patient Register (NPR) is managed by the National Board of Health and Welfare in Sweden and was established in 1964. In the beginning, it contained mainly information regarding patients in psychiatric care. From 1986 it included all inpatient care and from 2001 the NPR also covered specialist outpatient care in Sweden (125).

The information that can be retained from the NPR can be divided into 4 groups: patient data, geographical data, administrative data and medical data (ICD-diagnosis). The coverage from the NPR is approximately 99% for in-hospital care. For the specialist out-hospital care the same number is approximately 97% (126)

1.4.2 The Swedish Prescribed Drug Register

The Swedish Prescribed Drug Register is managed by the National Board of Health and Welfare in Sweden and was established in 2005. It contains all prescribed drugs which have been distributed at pharmacies. The prescribed drug is linked with a personal identification number used in Sweden. Drugs or medications used in hospitals or vaccines are not included in the register (127).

1.4.3 The National Cause of Death Register

The National Cause of Death Register is managed by the National Board of Health and Welfare in Sweden and was established in 1961. It contains information on the cause of death of all people deceased in Sweden except for stillborn babies (128). Coverage of the cause of death has been approximately 98-99% over the past few years (129).

1.4.4 SWEDEHEART

The Swedish Web system for Enhancement and Development of Evidence-based care in Heart Disease Evaluated According to Recommended Therapies (SWEDEHEART) was established in 2009 by merging 4 already existing registers, the Register of Information and Knowledge about Swedish Heart Intensive Care Admission (RIKS-HIA), the National Registry of Secondary Prevention (SEPHIA), the Swedish Coronary Angiography and Angioplasty Registry (SCAAR) and the Swedish Heart Surgery register.

(29)

Approximately 80,000 patients are enrolled in SWEDEHEART every year and the data collected consists of 106 variables including medical history, medical treatment before, during and after discharge, risk factors, demographics etc. SWEDEHEART has 100% coverage of patients undergoing heart surgery and angiography and approximately 60% of patients treated for an AMI. One of the reasons why SWEDHEART only has 60% coverage of AMI is that some patients with AMI are admitted to other wards than cardiology wards (130).

(30)

2 RESEARCH AIMS

The general aim was to use the glucose values in the OGTT as a prognostic tool in patients with ACS and newly discovered disturbances of the glucose metabolism and to evaluate the long-term prognosis in these patients in comparison to patients with NGT. Our specific aims were:

1) To assess cardiovascular outcome a decade after ACS and its relationship with repeated measurements of metabolic status and indices of glycemic abnormalities.

2) To study long-term prognosis in patients with AMI and glucose abnormalities and focus on the predictive value of OGTT test and HbA1c.

3) To study preoperative disturbances of glucose metabolism and mortality after CABG.

4) To study the prognostic importance of random prandial blood glucose tests and the risk of mortality and cardiovascular events in patients attending the ED.

(31)

3 MATERIALS AND METHODS

3.1 SUBJECTS

Study I. The study consisted of 1,062 patients, 281 women and 781 men, aged 32-80 years, admitted for ACS between 2006-2008 at the Department of Cardiology at Danderyd’s Hospital.

All patients without previously known diabetes underwent a standardised 75-g OGTT according to WHO criteria 4-5 days after admission. In 2010, 109 patients were identified as deceased, and 47 patients were lost to follow-up. The remaining 953 patients were asked to participate in a follow-up study whereupon 515 patients agreed to participate.

Study II. The study population consisted of 1,684 patients, aged 32-80 years, admitted for ACS between 2006-2013 at the Department of Cardiology at Danderyd’s Hospital. All patients without known diabetes underwent a standardised 75-g OGTT according to WHO criteria 4-5 days after admission and in some patients, an HbA1c value was controlled simultaneously. Of the 1,684 patients screened with OGTT, 841 (50%) had a registered HbA1c value too. An additional 433 patients had known diabetes.

Study III. 497 patients aged 40-86 years, without previous history of diabetes were recruited from Danderyd’s Hospital and Karolinska University Hospital. All patients underwent a first, isolated CABG at the Karolinska University Hospital between 2005-2013. Out of these 497 patients, 298 patients had an elective, non-emergent CABG during 2005-2008 and 199 patients were treated for an AMI at Danderyd’s Hospital and referred for a CABG at the time of their AMI during 2006 - 2013. All patients underwent a standardised 75-g OGTT according to WHO criteria prior to their surgery.

Study IV. All patients, aged 18-80 years, admitted to the ED with one (index at admission) random glucose level in four hospitals in Stockholm (Karolinska University Hospital Huddinge, Karolinska University Hospital Solna, Danderyd’s Hospital, Södersjukhuset) and at three hospitals in Gothenburg (Sahlgrenska Hospital, Östra Hospital and Mölndal Hospital) between 2008 to 2016 were be included. Patients with known diabetes were excluded and only patients with one random glucose measurement at the first visit to the ED was included during this time period.

3.2 METHODS – OUTCOMES, EXPLORING FACTORS AND REGISTERS

MI and unstable angina pectoris were defined according to the criteria recommended by the Joint European Society of Cardiology and the American College of Cardiology Committee (131).

Several registers were used to identify the patients and to find baseline information. The Swedish Web System for Enhancement and Development of Evidence-based Care in Heart Disease Evaluated According to Recommended Therapies Registry (SWEDEHEART), which is a national quality registry contains data on all patients with acute coronary syndrome including medical history, medical treatment before, during and after discharge, risk factors and demographics (130). Data was also collected from the national patient register which contains information on geographical data, ICD-diagnosis, patient data and administrative data (125). The Swedish Prescribed Drug Register contains information about all the prescribed

(32)

drugs which have been distributed at pharmacies. Drugs or medications used in hospitals or vaccines are not included in the register (127). The national cause of death register contains information on the cause of death of all people who have died in Sweden except for stillborn babies (128).

Ethical considerations: All studies complied with the Declaration of Helsinki and were approved by the regional research ethics committee in Stockholm. In study I, all participants received written and oral information regarding the study and gave their informed consent.

3.2.1 Study I-III

All participants without known diabetes underwent a 75-g OGTT according to WHO and were divided into four different glucose tolerance groups depending on the result of the OGTT (96).

Criteria for abnormalities of glucose metabolism:

• NGT: a fasting venous plasma glucose level < 6.1 mmol/l and a 2-h plasma glucose level at OGTT < 7.8 mmol/l

• IFG: a fasting venous plasma glucose level of ≥ 6.1 to < 7.0 mmol/l and a 2-h plasma glucose level at OGTT <7.8 mmol/l

• IGT: a fasting venous plasma glucose level < 6.1 mmol/l and a 2-h plasma glucose level at OGTT of ≥ 7.8 to <11.1 mmol/l

• T2DM: a fasting venous plasma glucose level ≥ 7.0 mmol/l, and/or a 2-h plasma glucose value at OGTT ≥ 11.1 mmol/l

Prediabetes is defined as IGT and/or IFG.

In study I-II, the OGTT was performed 4-5 days after the MI. In paper III, the OGTT was performed within 3 months of their CABG.

In study II, patients without known diabetes had an HbA1c checked at the time of AMI and were divided into three different glucose tolerance groups depending on the result of the HbA1c according to ADA criteria (38).

• NGT: HbA1c < 39 mmol/mol

• Prediabetes: HbA1c 39 - 47 mmol/mol

• T2DM: HbA1c ≥ 48 mmol/mol

3.2.2 Study IV

In study IV, a random blood glucose test was conducted upon admission to the ED. Depending on the result, patients were divided into four glucose tolerance groups according to ADA 2021 (38).

• Hypoglycaemia – Random plasma glucose value at < 3.9 mmol/l

• NGT - Random plasma glucose level at ≥ 3.9 to < 7.8 mmol/l

• Dysglycaemia - Random plasma glucose level ≥ 7.8 to <11.1 mmol/l.

• Hyperglycaemia - Random plasma glucose value ≥ 11.1 mmol/l

(33)

3.3 STUDY PROTOCOL 3.3.1 Study I

All patients completed an extensive questionnaire that included questions relating to weight, height, blood pressure, whether they smoke, use “snus”, do any physical activity, and which medications they take. BMI was calculated as the ratio of body weight in kilograms to the square of height in meters (kg/m²). All patients provided a blood sample for the analysis of HbA1c, FPG, fasting plasma insulin, lipids and creatinine.

The study patients were followed for 10 years regarding cardiovascular outcomes and all-cause mortality. Data regarding new MI, revascularization and CABG was collected from Swedeheart. Data regarding mortality and heart failure were collected from the patients’

medical records and by personal contacting patients when the data was missing.

Calculations for HbA1c, insulin resistance and β-cell function

Levels of HbA1c were determined with the Mono S method using high-performance liquid chromatography (Variant II; Bio-Rad Laboratories, Hercules, CA, USA) with a reference value of less than 5.2%.

Homeostatic Model Assessment (HOMA) was used to assess insulin sensitivity and beta-cell function, based on fasting insulin and glucose levels and according to published algorithms:

HOMA insulin resistance = (insulin pmol/L x glucose mmol/L)/22.5 and HOMA β-cell function =20 x insulin/(glucose-3.5) (16).

3.3.2 Study II

A retrospective study where all patients without known diabetes underwent an OGTT and for most patients, HbA1c values were measured simultaneously. Patient records were scrutinized and any missing information regarding OGTT and HbA1c was added. An HbA1c within 3 months of the ACS was accepted in those cases and HbA1c was not measured at the time of the ACS. All patients were followed up for all-cause mortality until December 25^th 2017.

Combined events were followed until December 31^th 2014.

Calculations for HbA1c

Levels of HbA1c were determined with the Mono S method using high-performance liquid chromatography (Variant II; Bio-Rad Laboratories, Hercules, CA, USA). In 2010 the international IFCC method was introduced in Sweden replacing the Mono S method (132).

3.3.3 Study III

A retrospective study. 199 patients were treated for an AMI at Danderyd University Hospital between 2006 and 2013 and were referred to Karolinska University Hospital for a CABG at the time of their AMI. 298 patients had an elective, non-emergent CABG performed at Karolinska University Hospital during 2005-2008.

Patients were identified, and baseline characteristics were collected from the SWEDEHEART register (130). Mortality information was obtained from the National Cause of Death Register.

Levels of HbA1c were determined with the Mono S method using high-performance liquid chromatography (Variant II; Bio-Rad Laboratories, Hercules, CA, USA).

(34)

3.3.4 Study IV

A retrospective study where all patients had their random blood glucose measured upon admission to the ED. At most hospitals, plasma blood glucose was measured, but at the hospitals in Gothenburg, venous blood glucose was measured instead. The formula “Plasma glucose = Venous blood glucose x 1.11” was used to equalize the measurements (133).

Information regarding medical history was collected from the National Patient Register.

Mortality information was collected from the National Cause of Death Register. Information about medication/drugs was collected from the Prescribed Drug Register. Information regarding socioeconomic status was collected from Statistiska centralbyrån (Statistics Sweden).

4 STATISTICAL ANALYSIS

Patient characteristics were described using frequencies and percentages for categorical variables and means and standard deviations (SD) for continuous variables. Statistical significance was set at p<0.05.

4.1 STUDY 1

To analyse differences in continuous variables between groups, the Kruskal-Wallis test followed by Mann-Whitney’s U-test, ANOVA and Chi-Square Test were performed. Data shows absolute risks.

4.2 STUDY II

Baseline characteristics are presented as medians with interquartile range for continuous and absolute numbers and percentages for categorical variables. Baseline characteristics stratified by glucose perturbation group were compared using the chi-square or the non-parametric Kruskal-Wallis test as applicable.

Time-to-event rates were calculated using survival analysis. Cox regression analysis was used to calculate Hazard Ratios (HR) and their 95% confidence intervals (CI) for all-cause

mortality. The proportionality of hazards assumption was not violated when tested using Schoenfeld residuals.

4.3 STUDY III

When comparing the three groups, the chi-squared test was used for categorical variables and ANOVA for continuous variables. The person-time in days contributed by each patient was calculated from the date of surgery to the date of death or the end of the follow-up (1 January 2018), whichever occurred first.

Cox regression was used to estimate the risk of all-cause mortality according to the following OGTT categories: NGT, prediabetes and diabetes, where NGT was the reference category.

Crude, age- and sex-adjusted, and multivariable-adjusted HRs and 95% CIs were calculated.

In the final multivariable model the following covariates were used: renal function (eGFR), left ventricular ejection fraction, body mass index, hypertension, pulmonary disease,

peripheral vascular disease, smoking, prior MI, prior percutaneous coronary intervention, and prior stroke.

Oral glucose tolerance test as a prognostic tool in patients with acute coronary syndrome

Karolinska Institutet, Department of Medicine Cardiology Unit, Karolinska University Hospital

Stockholm, Sweden

ORAL GLUCOSE TOLERANCE TEST AS A PROGNOSTIC TOOL IN PATIENTS WITH ACUTE CORONARY SYNDROME

Catarina Djupsjö

Stockholm 2022

Oral glucose tolerance test as a prognostic tool in patients with acute coronary syndrome

By Catarina Djupsjö

THESIS FOR DOCTORAL DEGREE (Ph.D.)

“The only true wisdom is in knowing you know nothing.”

ABSTRACT

LIST OF SCIENTIFIC PAPERS

CONTENTS

LIST OF ABBREVIATIONS

1 INTRODUCTION

2 RESEARCH AIMS

3 MATERIALS AND METHODS

4 STATISTICAL ANALYSIS