postprandial dysmetabolism and non-alcoholic fatty liver … · metabolism and the cvd mortality in...

Postprandial Dysmetabolism and Non-Alcoholic Fatty Liver Disease

in Relation to Type 2 Diabetes Mellitus and Cardiovascular Risk

The studies presented in this thesis were performed at the Diabetes Center of the

Department of Endocrinology in collaboration with the EMGO Institute and the

Department of Clinical Chemistry, VU University Medical Center, Amsterdam, The

Netherlands, and at the Diabetes Research Center in Hoorn, The Netherlands. Parts of

this work were supported by the Dutch Diabetes Research Foundation (grant no.

2001.00.052) and by funding from Novartis AG, Switzerland.

Financial support by the Dutch Diabetes Research Foundation for the publication of this

thesis is gratefully acknowledged.

Additional financial support for the publication of this thesis was kindly provided by:

Eli Lilly Nederland B.V., GlaxoSmithKline B.V., HemoCue Diagnostics B.V., Novartis

Pharma B.V., Novo Nordisk Farma B.V. and Pfizer B.V, Servier Nederland Farma B.V.

Cover

Cover illustration “Drieluik 2006” by Suzanne Kalf ([email protected]).

Printing

Ponsen & Looijen B.V., Wageningen, The Netherlands.

ISBN 978-90-6464-113-8

NUR 878

© Roger K. Schindhelm, The Netherlands.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval

system, or transmitted, in any form or by any means, electronic, photocopying, or

otherwise, without the permission of the author, or, when appropriate, of the

publishers of the publications.

VRIJE UNIVERSITEIT

Postprandial Dysmetabolism and Non-Alcoholic Fatty Liver Disease

in Relation to Type 2 Diabetes Mellitus and Cardiovascular Risk

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus

prof.dr. L.M. Bouter, in het openbaar te verdedigen

ten overstaan van de promotiecommissie van de faculteit der Geneeskunde

op maandag 21 mei 2007 om 10.45 uur in de aula van de universiteit,

De Boelelaan 1105

door

Roger Karlo Schindhelm

geboren te Kerkrade

promotor: prof.dr. R.J. Heine copromotoren: dr. M. Diamant dr. T. Teerlink

Financial support by the Netherlands Heart Foundation for the publication of this

thesis is gratefully acknowledged.

5

Voor mijn ouders

6

CONTENTS Chapter 1: General introduction and outline of the thesis 9

PART 1 Postprandial dysmetabolism, cardiovascular disease risk

and type 2 diabetes mellitus

Chapter 2: Determinants of postprandial triglyceride and glucose

responses following two consecutive fat-rich or carbohydrate-

rich meals in normoglycemic women and in women with type

2 diabetes: the Hoorn Prandial study.

Submitted

27

Chapter 3: Postprandial glucose and not triglyceride concentrations are

associated with carotid intima media thickness in women with

normal glucose metabolism: the Hoorn Prandial study

Atherosclerosis (in press)

45

Chapter 4: Fasting and postprandial glycoxidative and lipoxidative stress

are increased in women with type 2 diabetes mellitus

Submitted

63

Chapter 5: Comparison of the effect of two consecutive fat-rich and

carbohydrate-rich meals on levels of postprandial plasma

myeloperoxidase in women with and without type 2 diabetes

mellitus

Submitted

79

PART 2 Liver enzymes in relation to the metabolic syndrome, type

2 diabetes mellitus and cardiovascular disease risk

Chapter 6: Alanine aminotransferase as a marker of non-alcoholic fatty

liver disease in relation to type 2 diabetes mellitus and

cardiovascular disease (review).

Diabetes Metab Res Rev 2006;22:437-43.

95

Chapter 7:

Liver alanine aminotransferase, insulin resistance and

endothelial dysfunction in normotriglyceridemic subjects with

type 2 diabetes mellitus.

Eur J Clin Invest 2005;35:369-74.

113

7

Chapter 8: Alanine aminotransferase predicts coronary heart disease

events: a 10-year follow-up of the Hoorn study.

Atherosclerosis 2007;191;391-6.

129

Chapter 9.1: Alanine aminotransferase and the 6-years risk of the

metabolic syndrome in Caucasian men and women: the Hoorn

study.

Diabet Med 2007;24:430-5.

143

Chapter 9.2: No independent association of alanine aminotransferase with

risk of future type 2 diabetes in the Hoorn study (letter).

Diabetes Care 2005;28:2812.

155

Chapter 10.1:

Chapter 10.2:

General discussion - Methodological considerations

General discussion - Pathophysiological considerations

159

169

APPENDIX

Summary

188

Samenvatting voor niet ingewijden

191

Dankwoord

196

List of publications

200

About the author/Over de auteur 203

9

Chapter 1

GENERAL INTRODUCTION AND OUTLINE OF THE THESIS

CHAPTER 1

10

DIABETES MELLITUS AND THE METABOLIC SYNDROME

Diabetes mellitus is an endocrine disorder, characterized by a chronic elevation of

blood glucose levels (hyperglycemia). Type 1 diabetes mellitus (less than 10% of all

diabetes mellitus cases) is caused by an immune mediated destruction of the β-

cells resulting in an absolute insulin deficiency. This is the most prevalent form

among children and adolescents. In individuals with type 2 diabetes (DM2),

hyperglycemia is caused by a combination of an impairment of the beta-cells to

release sufficient insulin (β-cell dysfunction) and an impaired metabolic response of

peripheral tissues to circulating insulin (insulin resistance) [1,2]. Deficient action of

insulin on target tissues, either due to insulin resistance or insulin deficiency, not

only leads to hyperglycemia, but also to other metabolic abnormalities, including

abnormalities in fasting and postprandial lipid metabolism (dyslipidemia).

Dyslipidemia associated with DM2 is characterized by hypertriglyceridemia, low

high-density lipoprotein (HDL)-cholesterol, and small and dense low-density

lipoprotein (LDL) particles, which are more atherogenic than the more buoyant

forms [3-5].

In 1988, Reaven described the insulin resistance syndrome or syndrome X [6], now

called the metabolic syndrome. It was originally defined by the presence of

hyperinsulinemia (compensatory for the underlying insulin resistance), varying

degrees of glucose tolerance, hypertriglyceridemia and low plasma HDL cholesterol

concentration [6]. Remarkably, obesity or waist circumference was not part of the

original definition. In the most common working definition, as proposed by the

Adult Treatment Panel III of the National Cholesterol Education Program (NCEP) for

reasons of unification and clinical utility, the metabolic syndrome is characterized

by central obesity, hypertriglyceridemia, low plasma HDL-cholesterol, hypertension

and dysglycemia, i.e. impaired fasting glucose [7].

The prevalence of diabetes has increased substantially over the last decades. An

estimated 30 million people worldwide had diabetes in 1985. For the year 2000,

the World Health Organization has estimated the number of cases of adults with

diabetes to be 171 million and the total number is projected to rise to 366 million

by the year 2030 [8]. The most important factors contributing to this increase in

prevalence are ageing of the population and, in particular, the increasing

prevalence of obesity [9].

INTRODUCTION & OUTLINE

11

Among individuals with DM2, the cardiovascular disease (CVD) morbidity and

mortality is two to four times higher compared to individuals with normal glucose

metabolism and the CVD mortality in individuals with DM2 accounts for as much as

75% of all deaths [10,11]. In the Reykjavik Study, a population-based study with

over 18000 elderly participants, the relative impact of elevated glucose,

triglycerides and systolic blood pressure with regard to fatal and non-fatal CHD was

higher in women than in men [11]. In premenopausal women non-HDL cholesterol

concentrations are lower than in age-matched men, whereas the reverse is true

after menopause [12]. Indeed, postmenopausal women have higher fasting and

postprandial triglyceride levels [13], higher total cholesterol [14,15] and smaller

LDL particles [16], as compared to premenopausal women. This higher relative CVD

risk and the paucity of studies in postmenopausal women prompted us to perform

the postprandial studies described in this thesis in postmenopausal women.

DETERMINANTS OF CARDIOMETABOLIC RISK

New determinants of cardiometabolic risk

There are several conventional risk factors that can explain part of the excess CVD

risk in individuals with DM2. These include high LDL cholesterol, low HDL

cholesterol, hypertension, smoking, hyperglycemia, and high triglyceride

concentrations [17,18]. However, these risk factors, only explain part of the

increased CVD risk in individuals with DM2 [18]. For that reason additional risk

factors or determinants have been explored that might contribute to the excess

CVD risk in individuals with DM2. These include, but are not limited to,

postprandial dysmetabolism [19,20], endothelial dysfunction [21], low-grade

inflammation, non-alcoholic fatty liver disease [22], oxidative stress [23,24], and

advanced glycation end-products [25]. These additional potential determinants are

introduced in the next sections. A risk determinant is defined as a variable that is

directly or indirectly associated with outcome, regardless of the postulated

causality. In case of a risk factor, the association with outcome is presumed to be

causal and in case of a risk predictor, the variable predicts the outcome without

necessarily conveying causality. A risk marker reflects not only the presence of a

risk factor but also some form of susceptibility of the organism to the detrimental

effects of this risk factor. A risk marker is associated with the disease (statistically)

but need not be causally linked and it may be a measure of the disease process

itself [26].

CHAPTER 1

12

Postprandial glucose and triglycerides

Both fasting triglycerides and glucose are established risk factors of CVD in

individuals with and without DM2. A meta-regression analysis of 20 studies with

more than 95 000 people demonstrated a continuous relationship of plasma

glucose levels with incident CVD events [27], and a meta-analyses in over 57 000

individuals of 17 studies found that fasting triglyceride levels were associated with

CVD, also when adjusted for HDL-cholesterol [28]. However, a number of studies

suggest that postprandial or post-load glucose and postprandial triglyceride

concentrations are more strongly associated with some of the CVD risk factors and

mortality than the respective fasting values [29-31]. They debate the causal role of

postprandial glucose in relation to CVD [32], versus simply reflecting the

underlying metabolic abnormalities in for example lipid metabolism. It has also

been suggested that the relation between hyperglycemia and CVD is mediated by

oxidative stress [33]. Ceriello et al. showed a cumulative effect of postprandial

hyperglycemia and postprandial hypertriglyceridemia on oxidative stress

generation in subjects with DM2 [34]. Indeed, specific meal components, such as

fat and carbohydrates may exert acute effects on endothelial function [35], and

enhance the production of reactive oxygen species, which may also contribute to

CVD risk [36]. However, the relative contributions and mechanisms of postprandial

glucose levels and postprandial triglyceride levels with respect to atherosclerosis

and CVD risk are not yet clarified [32,37].

Endothelial functions and dysfunctions

The vascular endothelium is a large autocrine, paracrine and endocrine organ, and

plays an important role in the regulation of vascular homeostasis [21,38,39]. The

normal, healthy endothelium synthesizes and releases numerous substances in

response to a myriad of diverse chemical and physical stimuli. These factors are

important in maintaining and regulating vascular homeostasis, including vascular

tone, smooth muscle cell proliferation, inflammation, and coagulation and

fibrinolysis [40,41]. Endothelium derived nitric oxide (NO) is a key molecule in the

regulation of endothelial functions. NO is synthesized by a stereo-specific reaction

from L-arginine by the enzyme nitric oxide synthase (eNOS) [42,43]. Shear stress

increases the expression of eNOS, whereas asymmetrical dimethylarginine (ADMA)

is an inhibitor of eNOS [44]. The produced NO diffuses through the endothelial cell

membrane into the adjacent smooth muscle cells, and activates soluble guanylate

cyclase, that converts guanosine triphospate (GTP) to cyclic 3’5’-guanosine


13

monophophate (cGMP) [45], which ultimately decreases intracellular calcium levels

leading to relaxation of the smooth muscle cells and causing vasodilatation. In

addition, NO also inhibits platelet adherence and aggregation, leukocyte adhesion

and proliferation of smooth muscle cells [46]. A disturbance of the endothelial

regulatory mechanisms may lead to endothelial dysfunctioning, which can be

defined as the loss of one or more of these functions [47]. Dysfunctioning of the

vascular endothelium is recognized to be an early step in the development of

atherosclerosis [48-51]. Endothelial function is difficult to assess. An optimal

methodology for studying endothelial function does not exist. Over the last

decades, multiple biochemical markers and invasive and non-invasive tests have

been introduced.

A number of biochemical markers have been identified that reflect endothelial

(dys)function, including soluble adhesion molecules, E-selectin and von Willebrand

factor (vWF). The soluble adhesion molecules, intercellular adhesion molecule 1

(ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) facilitate endothelial

adhesion of circulation leukocytes. These molecules are not endothelium specific;

also other cell types including smooth muscle cells and monocytes express them.

The expression of ICAM-1 and VCAM-1 is increased in response to inflammatory

cytokines [52]. Higher plasma concentrations of VCAM-1 and ICAM-1 have been

reported in DM2 [53,54]. E-selectin is specifically expressed by stimulated

endothelial cells and exposed on their surface. The soluble form of E-selectin may

result from shedding of damaged or activated endothelial cells [55]. High plasma

levels of E-selectin have been observed in obese individuals [56], in hypertensive

patients [57] and in individuals with DM2 [58]. Von Willebrand factor (vWF) is an

endothelial ligand for platelet glycoproteins and is endothelium specific. In case of

endothelial cell injury vWF is released. Major CVD risk factors, including smoking

[59], hypertension [60] and DM2 [61] are associated with increased plasma levels of

vWF. In addition, vWF is a predictor of CVD morbidity and mortality [62,63].

A non-invasive method for measuring endothelial function using high-resolution

vascular ultrasound was first described by Celermajer and colleagues [64]. In this

procedure, the arterial diameter of the brachial artery is subsequently measured in

rest, during reactive hyperemia (leading to flow mediated vasodilatation (FMD), an

endothelium-dependent response) and following sublingual administration of

nitroglycerin (an endothelium independent, nitroglycerin mediated dilatation,

NMD). The responses during reactive hyperemia are measured after a temporary

(usually 4 to 5 minutes) occlusion of the brachial artery using a cuff placed on the

CHAPTER 1

14

forearm [65,66]. FMD is expressed as (percentual) change in diameter after the

stimulus. A higher FMD reflects a “better” endothelial function. [67]. One of the

principal mediators of the FMD response is endothelium-derived NO [68]. The

brachial FMD is closely correlated to coronary endothelial function [46], and

impaired FMD predicts future cardiovascular events [69,70].

Oxidative stress and advanced glycation end-products

Numerous studies have identified hyperglycemia as an important risk factor in the

development of diabetic complications (reviewed in [71]), and several mechanisms

have been proposed [72,73]. These included, but were not limited to altered

lipoprotein metabolism, oxidative stress, dicarbonyl stress, and glycation reactions.

Glycation is the non-enzymatic reaction of glucose and other reducing sugars with

amino groups of proteins. The amino groups of the side chains of arginine and

lysine are the primary targets for this type of posttranslational modification. Over

time, the initial glycation products undergo intramolecular rearrangements and

oxidation reactions (glycoxidation) and ultimately transform into stable so-called

advanced glycation end-products (AGEs) (Figure 1.1). AGE-modification of proteins

can alter or limit their functional or structural properties, which ultimately can lead

to tissue damage as seen in age-related pathologies and diabetes. AGE formation is

promoted by high concentrations of sugars, including glucose, and α-oxoaldehydes

(e.g. 3-deoxyglucosone, methylglyoxal and glyoxal) and enhanced by oxidative

stress [74]. The mechanisms linking hyperglycemia to micro- and macrovascular

complications overlap and interact with each other, e.g. AGE formation and

carbonyl stress may lead to oxidative stress and oxidative stress may enhance AGE

formation [75].

Low-grade inflammation

The insulin resistant state is accompanied by (systemic) low-grade inflammation.

The proinflammatory cytokines include interleukin-6 (IL-6) and tumor necrosis

factor-α (TNF-α). They induce attraction of inflammatory cells and the production of

acute-phase proteins, including C-reactive protein (CRP), by the liver. IL-6 is a

pleiotropic cytokine, required to control the extent of local and systemic

inflammatory responses [76]. IL-6 is increased in hypertension [77], obesity [78],

DM2 [79], and non-alcoholic fatty liver disease (NAFLD) [80]. IL-6 is positively

associated with CVD risk in individuals without DM2 [81].


15

glucose

Schiff base

Amadori adduct

glucosone

methyglyoxal

glyoxal

glycolaldehyde

methyglyoxal

glyoxal

deoxyglucosones

AGEs and

glycoxidation

products

[O2]

[ONOO-]

[O2]

[O2]

[ONOO-]

[O2]

[O2]

[O2]

glucose

Schiff base

Amadori adduct

glucosone

methyglyoxal

glyoxal

glycolaldehyde

methyglyoxal

glyoxal

deoxyglucosones

AGEs and

glycoxidation

products

[O2]

[ONOO-]

[O2]

[O2]

[ONOO-]

[O2]

[O2]

[O2]

Figure 1.1. Pathways for the formation of advanced glycation end-products (AGE) and AGE-

precursors (adapted from: Thorpe SR, Baynes JW. Amino Acids 2003;25:275-281.)

C-reactive protein (CRP) is an acute phase protein, which is produced by the liver

and is widely used as an indicator of low-grade inflammation. The synthesis of CRP

is largely mediated by cytokines, in particular by IL-6. CVD risk factors that increase

CRP concentrations include obesity [82], hypertension [77], dyslipidemia [83], and

DM2 [83,84]. Some [85], but not all studies [80], have demonstrated an association

of CRP with NAFLD. CRP concentrations are higher in individuals with CVD as

compared to those without CVD. In addition, CRP is positively associated with risk

of CVD, both in patients with DM2 [86,87] and in individuals without DM2 [88]. A

more classical and clinical marker of (low-grade) inflammation is the blood

leukocyte count, which is also associated with CVD events [89,90]. The exact role

of leukocytes in the pathogenesis of CVD is not yet fully elucidated. Postprandial

recruitment and activation of these cells have been suggested to contribute to

endothelial dysfunction [91]. Myeloperoxidase (MPO) is a heme protein which is

abundantly expressed in leukocytes, and released upon activation into phagocytic

vacuoles and in the extra cellular space [92,93], and is regarded as maker of low-

grade systemic inflammation [94]. Several in vitro and in vivo studies have shown

that MPO may interact with the vessel wall by various mechanisms including

binding and transcytosis through endothelial cells, and inflicts damage by

hypochlorous acid generation, nitric oxide oxidation and tyrosine nitration [95].

MPO has been associated with recurrent coronary events [96], and endothelial

dysfunction [97].

CHAPTER 1

16

Non-alcoholic fatty liver disease and liver enzymes

NAFLD includes a wide spectrum of liver pathology, ranging from fatty liver alone

to the more severe non-alcoholic steatohepatitis. It usually has a benign clinical

course, but it may progress to steatohepatitis, fibrosis, and cirrhosis and ultimately

to liver failure. NAFLD resembles alcohol-induced liver disease, but develops in

subjects who are not heavy alcohol consumers and have negative tests for viral and

autoimmune liver diseases [98-102]. This condition was first described in detail by

Ludwig and colleagues in 1980, who reported 20 moderately obese patients with

liver biopsy changes resembling alcohol induced hepatitis, although none of these

patients reported a history of alcohol abuse [103]. Only in recent years, NAFLD has

gained appreciation as a pathogenic factor of insulin resistance and DM2. In

addition, several studies have shown an association between NAFLD and features of

the metabolic syndrome, including dyslipidemia and (visceral) obesity, stressing the

association with insulin resistance as an important feature of NAFLD. Currently,

NAFLD is considered by some authors to be the hepatic component of the

metabolic syndrome [104,105]. In the majority of cases, NAFLD causes

asymptomatic elevation of liver enzyme levels (including alanine aminotransferase

[ALT], aspartate aminotransferase and γ-glutamyltransferase) [106]. Of these liver

enzymes, ALT is most closely related to liver fat accumulation [107], and

consequently ALT has been used as a marker of NAFLD.

The exact mechanisms leading to hepatic triglyceride accumulation are not

completely understood. Insulin resistance is an early important mechanism leading

to an increased free fatty acid (FFA) influx into the liver [108,109], contributing to

hepatic triglyceride accumulation. In addition, the hepatic FFA pool is increased by

de novo lipogenesis [110]. The quantitative importance of the contribution of the

de novo lipogenesis to the hepatic fat pool is disputed. Hyperinsulinemia in

combination with a high FFA flux and hyperglycemia are known to up-regulate

lipogenic transcription factors. In addition, pathways that decrease the hepatic FFA

pool, i.e. both FFA oxidation and efflux of lipids from the liver are impaired. The

increased availability of FFA, glucose and insulin contribute to the increase of

malonyl-CoA by stimulating acetyl-CoA carboxylase that converts acetyl-CoA to

malonyl-CoA. Malonyl-CoA is an inhibitor of carnitine palmitoyl transferase 1 (CPT-

1), required to transport FFA into the mitochondria. The inhibition of CPT-1

prevents FFA oxidation and facilitates accumulation and re-esterification of long-

chain fatty-acyl CoA and subsequent formation of triglycerides [111].


17

Evidence is accumulating that patients with NAFLD have an increased risk to

develop CVD, but the mechanisms linking NAFLD to CVD are not yet fully

understood.

HOW DO THE VARIOUS CARDIOMETABOLIC RISK DETERMINANTS

RELATE TO EACH OTHER?

The interrelationship of the above mentioned cardiometabolic determinants with

respect to CVD risk is complex. In particular it is not yet possible to infer any

causality for any of these new risk markers. Indeed, NAFLD may be a cause or a

consequence of the metabolic syndrome and postprandial dysmetabolism may

contribute to the development and progression of NAFLD, but also, once a fatty

liver has evolved, postprandial dysmetabolism may aggravate, especially in patients

with DM2. These cardiometabolic determinants are likely to interact and contribute

to the enhanced CVD risk in patients with DM2 and/or individuals with the

metabolic syndrome. A better understanding of these interrelationships and their

relative contribution to CVD is of importance to design strategies, which will lower

CVD risk in populations at risk.

AIMS AND OUTLINE OF THE THESIS

This thesis is divided into two parts.

Part 1 (Chapters 2 to 5) presents the results of studies assessing the responses to

two consecutive fat-rich or carbohydrate-rich meals on the (apo)lipoproteins and

glucose and to study mechanisms that link postprandial glucose and triglyceride

levels to CVD risk in patients with DM2 and individuals with normal glucose

metabolism.

Part 2 (Chapters 6 to 9) focuses on ALT as a marker for NAFLD in relation to CVD,

the metabolic syndrome and DM2.

CHAPTER 1

18

REFERENCES

1. Kahn SE. The relative contributions of insulin resistance and beta-cell dysfunction to

the pathophysiology of Type 2 diabetes. Diabetologia 2003;46:3-19.

2. Weir GC, Bonner-Weir S. Five Stages of Evolving Beta-Cell Dysfunction During

Progression to Diabetes. Diabetes 2004;53:S16-S21.

3. Betteridge DJ. Diabetic dyslipidaemia. Eur J Clin Invest 1999;29 Suppl 2:12-6.

4. Taskinen MR. Diabetic dyslipidemia. Atheroscler Suppl 2002;3:47-51.

5. Patel J. Dyslipidaemia in diabetes. Clin Evid 2006;555-75.

6. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease.

Diabetes 1988;37:1595-607.

7. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in

Adults. 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;285:2486-

97.

8. Wild S, Roglic G, Green A, Sicree R, King H. Global Prevalence of Diabetes: Estimates

for the year 2000 and projections for 2030. Diabetes Care 2004;27:1047-53.

9. Kopelman PG. Obesity as a medical problem. Nature 2000;404:635-43.

10. Becker A, Bos G, de Vegt F, et al. Cardiovascular events in type 2 diabetes:

comparison with nondiabetic individuals without and with prior cardiovascular

disease. 10-year follow-up of the Hoorn Study. Eur Heart J 2003;24:1406-13.

11. Jonsdottir LS, Sigfusson N, Gudnason V, Sigvaldason H, Thorgeirsson G. Do lipids,

blood pressure, diabetes, and smoking confer equal risk of myocardial infarction in

women as in men? The Reykjavik Study. J Cardiovasc Risk 2002;9:67-76.

12. Gardner CD, Winkleby MA, Fortmann SP. Population frequency distribution of non-

high-density lipoprotein cholesterol (Third National Health and Nutrition

Examination Survey [NHANES III], 1988-1994). Am J Cardiol 2000;86:299-304.

13. van Beek AP, Ruijter-Heijstek FC, Erkelens DW, de Bruin TW. Menopause is

associated with reduced protection from postprandial lipemia. Arterioscler Thromb

Vasc Biol 1999;19:2737-41.

14. Akahoshi M, Soda M, Nakashima E, et al. Effects of menopause on trends of serum

cholesterol, blood pressure, and body mass index. Circulation 1996;94:61-6.

15. Hjortland MC, McNamara PM, Kannel WB. Some atherogenic concomitants of

menopause: The Framingham Study. Am J Epidemiol 1976;103:304-11.

16. Wakatsuki A, Ikenoue N, Shinohara K, Watanabe K, Fukaya T. Small low-density

lipoprotein particles and endothelium-dependent vasodilation in postmenopausal

women. Atherosclerosis 2004;177:329-36.

17. Turner RC, Millns H, Neil HA, et al. Risk factors for coronary artery disease in non-

insulin dependent diabetes mellitus: United Kingdom Prospective Diabetes Study

(UKPDS: 23). BMJ 1998;316:823-8.


19

18. Niskanen L, Turpeinen A, Penttila I, Uusitupa MI. Hyperglycemia and compositional

lipoprotein abnormalities as predictors of cardiovascular mortality in type 2

diabetes: a 15-year follow-up from the time of diagnosis. Diabetes Care

1998;21:1861-9.

19. Zilversmit DB. Atherogenesis: a postprandial phenomenon. Circulation

1979;60:473-85.

20. Tushuizen ME, Diamant M, Heine RJ. Postprandial dysmetabolism and

cardiovascular disease in type 2 diabetes. Postgrad Med J 2005;81:1-6.

21. Lerman A, Zeiher AM. Endothelial function: cardiac events. Circulation

2005;111:363-8.

22. Targher G. Non-alcoholic fatty liver disease as a determinant of cardiovascular

disease. Atherosclerosis 2006.

23. Heitzer T, Schlinzig T, Krohn K, Meinertz T, Munzel T. Endothelial Dysfunction,

Oxidative Stress, and Risk of Cardiovascular Events in Patients With Coronary Artery

Disease. Circulation 2001;104:2673-8.

24. Kuroki T, Isshiki K, King GL. Oxidative Stress: The Lead or Supporting Actor in the

Pathogenesis of Diabetic Complications. J Am Soc Nephrol 2003;14:S216-S220.

25. Kilhovd BK, Berg TJ, Birkeland KI, Thorsby P, Hanssen KF. Serum levels of advanced

glycation end products are increased in patients with type 2 diabetes and coronary

heart disease. Diabetes Care 1999;22:1543-8.

26. Vasan RS. Biomarkers of cardiovascular disease: molecular basis and practical

considerations. Circulation 2006;113:2335-62.

27. Coutinho M, Gerstein HC, Wang Y, Yusuf S. The relationship between glucose and

incident cardiovascular events. A metaregression analysis of published data from

20 studies of 95,783 individuals followed for 12.4 years. Diabetes Care

1999;22:233-40.

28. Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk

factor. Am J Cardiol 1998;81:7B-12B.

29. Teno S, Uto Y, Nagashima H, et al. Association of postprandial hypertriglyceridemia

and carotid intima-media thickness in patients with type 2 diabetes. Diabetes Care

2000;23:1401-6.

30. de Vegt F, Dekker JM, Ruhe HG, et al. Hyperglycaemia is associated with all-cause

and cardiovascular mortality in the Hoorn population: the Hoorn Study. Diabetologia

1999;42:926-31.

31. 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 Diagnostic

criteria in Europe Lancet 1999;354:617-21.

32. Heine RJ, Dekker JM. Beyond postprandial hyperglycaemia: metabolic factors

associated with cardiovascular disease. Diabetologia 2002;45:461-75.

CHAPTER 1

20

33. Ceriello A. The post-prandial state and cardiovascular disease: relevance to diabetes

mellitus. Diabetes Metab Res Rev 2000;16:125-32.

34. Ceriello A, Taboga C, Tonutti L, et al. Evidence for an independent and cumulative

effect of postprandial hypertriglyceridemia and hyperglycemia on endothelial

dysfunction and oxidative stress generation: effects of short- and long-term

simvastatin treatment. Circulation 2002;106:1211-8.

35. de Koning EJ, Rabelink TJ. Endothelial function in the post-prandial state.

Atheroscler Suppl 2002;3:11-6.

36. Tushuizen ME, Nieuwland R, Scheffer PG, et al. Two consecutive high-fat meals

affect endothelial-dependent vasodilation, oxidative stress and cellular

microparticles in healthy men. J Thromb Haemost 2006;4:1003-10.

37. Heine RJ, Balkau B, Ceriello A et al. What does postprandial hyperglycaemia mean?

Diabet Med 2004;21:208-13.

38. Vanhoutte PM. Endothelium and control of vascular function. State of the Art

lecture. Hypertension 1989;13:658-67.

39. Jaffe EA. Cell biology of endothelial cells. Hum Pathol 1987;18:234-9.

40. Braun M, Pietsch P, Schror K, Baumann G, Felix SB. Cellular adhesion molecules on

vascular smooth muscle cells. Cardiovasc Res 1999;41:395-401.

41. Gray GA, Battistini B, Webb DJ. Endothelins are potent vasoconstrictors, and much

more besides. Trends Pharmacol Sci 2000;21:38-40.

42. Palmer RM, Ashton DS, Moncada S. Vascular endothelial cells synthesize nitric oxide

from L-arginine. Nature 1988;333:664-6.

43. Moncada S, Higgs A. The L-arginine-nitric oxide pathway. N Engl J Med

1993;329:2002-12.

44. Lin KY, Ito A, Asagami T, et al. Impaired nitric oxide synthase pathway in diabetes

mellitus: role of asymmetric dimethylarginine and dimethylarginine

dimethylaminohydrolase. Circulation 2002;106:987-92.

45. Zhao Y, Brandish PE, Ballou DP, Marletta MA. A molecular basis for nitric oxide

sensing by soluble guanylate cyclase. Proc Natl Acad Sci U S A 1999;96:14753-8.

46. Anderson TJ, Uehata A, Gerhard MD et al. Close relation of endothelial function in

the human coronary and peripheral circulations. J Am Coll Cardiol 1995;26:1235-

41.

47. Davignon J, Ganz P. Role of Endothelial Dysfunction in Atherosclerosis. Circulation

2004;109:III-27.

48. Ross R. The pathogenesis of atherosclerosis--an update. N Engl J Med

1986;314:488-500.

49. Celermajer DS. Endothelial dysfunction: does it matter? Is it reversible? J Am Coll

Cardiol 1997;30:325-33.

50. Toborek M, Kaiser S. Endothelial cell functions. Relationship to atherogenesis. Basic

Res Cardiol 1999;94:295-314.

51. Lusis AJ. Atherosclerosis. Nature 2000;407:233-41.


21

52. Pigott R, Dillon LP, Hemingway IH, Gearing AJ. Soluble forms of E-selectin, ICAM-1

and VCAM-1 are present in the supernatants of cytokine activated cultured

endothelial cells. Biochem Biophys Res Commun 1992;187:584-9.

53. Matsumoto K, Sera Y, Nakamura H, Ueki Y, Miyake S. Serum concentrations of

soluble adhesion molecules are related to degree of hyperglycemia and insulin

resistance in patients with type 2 diabetes mellitus. Diabetes Res Clin Pract

2002;55:131-8.

54. Takeuchi N, Kawamura T, Kanai A, et al. The effect of cigarette smoking on soluble

adhesion molecules in middle-aged patients with Type 2 diabetes mellitus. Diabet

Med 2002;19:57-64.

55. Leeuwenberg JF, Smeets EF, Neefjes JJ, et al. E-selectin and intercellular adhesion

molecule-1 are released by activated human endothelial cells in vitro. Immunology

1992;77:543-9.

56. Matsumoto K, Sera Y, Abe Y, et al. High serum concentrations of soluble E-selectin

correlate with obesity but not fat distribution in patients with type 2 diabetes

mellitus. Metabolism 2002;51:932-4.

57. Blann AD, Tse W, Maxwell SJ, Waite MA. Increased levels of the soluble adhesion

molecule E-selectin in essential hypertension. J Hypertens 1994;12:925-8.

58. Matsumoto K, Sera Y, Abe Y, Ueki Y, Miyake S. Serum Concentrations of Soluble

Vascular Cell Adhesion Molecule-1 and E-Selectin Are Elevated in Insulin-Resistant

Patients With Type 2 Diabetes. Diabetes Care 2001;24:1697-8.

59. Blann AD, Bushell D, Davies A, et al. von Willebrand factor, the endothelium and

obesity. Int J Obes Relat Metab Disord 1993;17:723-5.

60. Blann AD, Naqvi T, Waite M, McCollum CN. von Willebrand factor and endothelial

damage in essential hypertension. J Hum Hypertens 1993;7:107-11.

61. Ibrahim HA, El Meligi AA, Abdel-Hamid M, Elhendy A. Relations between von

Willebrand factor, markers of oxidative stress and microalbuminuria in patients with

type 2 diabetes mellitus. Med Sci Monit 2004;10:CR85-CR89.

62. Rumley A, Lowe GD, Sweetnam PM, Yarnell JW, Ford RP. Factor VIII, von Willebrand

factor and the risk of major ischaemic heart disease in the Caerphilly Heart Study.

Br J Haematol 1999;105:110-6.

63. Jager A, van Hinsbergh VW, Kostense PJ, et al. von Willebrand factor, C-reactive

protein, and 5-year mortality in diabetic and nondiabetic subjects: the Hoorn Study.

Arterioscler Thromb Vasc Biol 1999;19:3071-8.

64. Celermajer DS, Sorensen KE, Gooch VM, et al. Non-invasive detection of endothelial

dysfunction in children and adults at risk of atherosclerosis. Lancet 1992;340:1111-

5.

65. Corretti MC, Plotnick GD, Vogel RA. Technical aspects of evaluating brachial artery

vasodilatation using high-frequency ultrasound. Am J Physiol 1995;268:H1397-

H1404.

CHAPTER 1

22

66. Corretti MC, Anderson TJ, Benjamin EJ, et al. Guidelines for the ultrasound

assessment of endothelial-dependent flow-mediated vasodilation of the brachial

artery: a report of the International Brachial Artery Reactivity Task Force. J Am Coll

Cardiol 2002;39:257-65.

67. Henry RM, Ferreira I, Kostense PJ et al. Type 2 diabetes is associated with impaired

endothelium-dependent, flow-mediated dilation, but impaired glucose metabolism

is not; The Hoorn Study. Atherosclerosis 2004;174:49-56.

68. Joannides R, Haefeli WE, Linder L, et al. Nitric oxide is responsible for flow-

dependent dilatation of human peripheral conduit arteries in vivo. Circulation

1995;91:1314-9.

69. Neunteufl T, Katzenschlager R, Hassan A, et al. Systemic endothelial dysfunction is

related to the extent and severity of coronary artery disease. Atherosclerosis

1997;129:111-8.

70. Gokce N, Keaney JF, Jr., Hunter LM, et al. Risk Stratification for Postoperative

Cardiovascular Events via Noninvasive Assessment of Endothelial Function: A

Prospective Study. Circulation 2002;105:1567-72.

71. Stettler C, Allemann S, Juni P, et al. Glycemic control and macrovascular disease in

types 1 and 2 diabetes mellitus: Meta-analysis of randomized trials. Am Heart J

2006;152:27-38.

72. Peppa M, Uribarri J, Vlassara H. Advanced glycoxidation. A new risk factor for

cardiovascular disease? Cardiovasc Toxicol 2002;2:275-87.

73. Brownlee M. The pathobiology of diabetic complications: a unifying mechanism.

Diabetes 2005;54:1615-25.

74. Beisswenger PJ, Howell SK, Nelson RG, Mauer M, Szwergold BS. Alpha-oxoaldehyde

metabolism and diabetic complications. Biochem Soc Trans 2003;31:1358-63.

75. Chung SSM, Ho ECM, Lam KSL, Chung SK. Contribution of Polyol Pathway to

Diabetes-Induced Oxidative Stress. J Am Soc Nephrol 2003;14:S233-S236.

76. Hirano T, Akira S, Taga T, Kishimoto T. Biological and clinical aspects of interleukin

6. Immunol Today 1990;11:443-9.

77. Bautista LE, Vera LM, Arenas IA, Gamarra G. Independent association between

inflammatory markers (C-reactive protein, interleukin-6, and TNF-alpha) and

essential hypertension. J Hum Hypertens 2005;19:149-54.

78. Park HS, Park JY, Yu R. Relationship of obesity and visceral adiposity with serum

concentrations of CRP, TNF-alpha and IL-6. Diabetes Res Clin Pract 2005;69:29-35.

79. Tuttle HA, Davis-Gorman G, Goldman S, Copeland JG, McDonagh PF.

Proinflammatory cytokines are increased in type 2 diabetic women with

cardiovascular disease. J Diabetes Complications 2004;18:343-51.

80. Haukeland JW, Damas JK, Konopski Z, et al. Systemic inflammation in nonalcoholic

fatty liver disease is characterized by elevated levels of CCL2. J Hepatol

2006;44:1167-74.


23

81. Cesari M, Penninx BW, Newman AB, et al. Inflammatory markers and onset of

cardiovascular events: results from the Health ABC study. Circulation

2003;108:2317-22.

82. Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW. C-reactive protein in healthy

subjects: associations with obesity, insulin resistance, and endothelial dysfunction:

a potential role for cytokines originating from adipose tissue? Arterioscler Thromb

Vasc Biol 1999;19:972-8.

83. Ridker PM, Buring JE, Cook NR, Rifai N. C-reactive protein, the metabolic syndrome,

and risk of incident cardiovascular events: an 8-year follow-up of 14 719 initially

healthy American women. Circulation 2003;107:391-7.

84. Stehouwer CD, Gall MA, Twisk JW, et al. Increased urinary albumin excretion,

endothelial dysfunction, and chronic low-grade inflammation in type 2 diabetes:

progressive, interrelated, and independently associated with risk of death. Diabetes

2002;51:1157-65.

85. Kerner A, Avizohar O, Sella R, et al. Association between elevated liver enzymes and

C-reactive protein: possible hepatic contribution to systemic inflammation in the

metabolic syndrome. Arterioscler Thromb Vasc Biol 2005;25:193-7.

86. Schulze MB, Rimm EB, Li T, et al. C-reactive protein and incident cardiovascular

events among men with diabetes. Diabetes Care 2004;27:889-94.

87. Diamant M, Lamb HJ, van de Ree MA, et al. The association between abdominal

visceral fat and carotid stiffness is mediated by circulating inflammatory markers in

uncomplicated type 2 diabetes. J Clin Endocrinol Metab 2005;90:1495-501.

88. Koenig W, Sund M, Frohlich M, et al. C-Reactive protein, a sensitive marker of

inflammation, predicts future risk of coronary heart disease in initially healthy

middle-aged men: results from the MONICA (Monitoring Trends and Determinants

in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation

1999;99:237-42.

89. Friedman GD, Klatsky AL, Siegelaub AB. The leukocyte count as a predictor of

myocardial infarction. N Engl J Med 1974;290:1275-8.

90. Phillips AN, Neaton JD, Cook DG, Grimm RH, Shaper AG. Leukocyte count and risk

of major coronary heart disease events. Am J Epidemiol 1992;136:59-70.

91. Van Oostrom AJ, Sijmonsma TP, Verseyden C et al. Postprandial recruitment of

neutrophils may contribute to endothelial dysfunction. J Lipid Res 2003;44:576-83.

92. Nauseef WM. Insights into myeloperoxidase biosynthesis from its inherited

deficiency. J Mol Med 1998;76:661-8.

93. Winterbourn CC. Comparative reactivities of various biological compounds with

myeloperoxidase-hydrogen peroxide-chloride, and similarity of the oxidant to

hypochlorite. Biochim Biophys Acta 1985;840:204-10.

94. Andelid K, Bake B, Rak S, et al. Myeloperoxidase as a marker of increasing systemic

inflammation in smokers without severe airway symptoms. Respir Med (in press)

2006.

CHAPTER 1

24

95. Daugherty A, Dunn JL, Rateri DL, Heinecke JW. Myeloperoxidase, a catalyst for

lipoprotein oxidation, is expressed in human atherosclerotic lesions. J Clin Invest

1994;94:437-44.

96. Baldus S, Heeschen C, Meinertz T, et al. Myeloperoxidase serum levels predict risk

in patients with acute coronary syndromes. Circulation 2003;108:1440-5.

97. Vita JA, Brennan ML, Gokce N, et al. Serum myeloperoxidase levels independently

predict endothelial dysfunction in humans. Circulation 2004;110:1134-9.

98. Angulo P. Nonalcoholic Fatty Liver Disease. N Engl J Med 2002;346:1221-31.

99. Matteoni CA, Younossi ZM, Gramlich T, et al. Nonalcoholic fatty liver disease: a

spectrum of clinical and pathological severity. Gastroenterology 1999;116:1413-9.

100. Brunt EM. Nonalcoholic steatohepatitis: definition and pathology. Semin Liver Dis

2001;21:3-16.

101. Falck-Ytter Y, Younossi ZM, Marchesini G, McCullough AJ. Clinical features and

natural history of nonalcoholic steatosis syndromes. Semin Liver Dis 2001;21:17-

26.

102. Brunt EM, Neuschwander-Tetri BA, Oliver D, Wehmeier KR, Bacon BR. Nonalcoholic

steatohepatitis: histologic features and clinical correlations with 30 blinded biopsy

specimens. Hum Pathol 2004;35:1070-82.

103. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic

experiences with a hitherto unnamed disease. Mayo Clin Proc 1980;55:434-8.

104. Cortez-Pinto H, Camilo ME, Baptista A, De Oliveira AG, De Moura MC. Non-alcoholic

fatty liver: another feature of the metabolic syndrome? Clin Nutr 1999;18:353-8.

105. Marchesini G, Brizi M, Bianchi G, et al. Nonalcoholic fatty liver disease: a feature of

the metabolic syndrome. Diabetes 2001;50:1844-50.

106. Clark JM, Brancati FL, Diehl AM. The prevalence and etiology of elevated

aminotransferase levels in the United States. Am J Gastroenterol 2003;98:960-7.

107. Westerbacka J, Corner A, Tiikkainen M, et al. Women and men have similar amounts

of liver and intra-abdominal fat, despite more subcutaneous fat in women:

implications for sex differences in markers of cardiovascular risk. Diabetologia

2004;47:1360-9.

108. Skowronski R, Hollenbeck CB, Varasteh BB, Chen YD, Reaven GM. Regulation of non-

esterified fatty acid and glycerol concentration by insulin in normal individuals and

patients with type 2 diabetes. Diabet Med 1991;8:330-3.

109. Coppack SW, Jensen MD, Miles JM. In vivo regulation of lipolysis in humans. J Lipid

Res 1994;35:177-93.

110. Diraison F, Moulin P, Beylot M. Contribution of hepatic de novo lipogenesis and

reesterification of plasma non esterified fatty acids to plasma triglyceride synthesis

during non-alcoholic fatty liver disease. Diabetes Metab 2003;29:478-85.

111. Lam TK, Carpentier A, Lewis GF, et al. Mechanisms of the free fatty acid-induced

increase in hepatic glucose production. Am J Physiol Endocrinol Metab

2003;284:E863-E873.

25

Part 1

POSTPRANDIAL DYSMETABOLISM, CARDIOVASCULAR DISEASE RISK AND TYPE 2 DIABETES MELLITUS

27

Chapter 2

DETERMINANTS OF POSTPRANDIAL TRIGLYCERIDE AND GLUCOSE RESPONSES FOLLOWING TWO CONSECUTIVE FAT-RICH OR

CARBOHYDRATE-RICH MEALS IN NORMOGLYCEMIC WOMEN AND IN

WOMEN WITH TYPE 2 DIABETES. THE HOORN PRANDIAL STUDY.

Marjan Alssema1*

Roger K. Schindhelm2*

Jacqueline M. Dekker1

Michaela Diamant2

Giel Nijpels1,3

Tom Teerlink4

Peter G. Scheffer4

Piet J. Kostense1,5

Robert J. Heine1,2

From the 1EMGO Institute, and the Departments of 2Endocrinology/Diabetes Center,

3General Practice, 4Clinical Chemistry and 5Clinical Epidemiology and Biostatistics,

VU University Medical Center, Amsterdam, The Netherlands

*Both authors contributed equally

Submitted

CHAPTER 2

28

ABSTRACT

Background: Both postprandial hyperglycemia and hypertriglyceridemia have been

identified as risk markers for cardiovascular disease. However, the determinants of

the magnitude of postprandial triglyceride and glucose responses are largely

unknown. The objective was to assess potential determinants of postprandial

glucose and triglyceride responses following two consecutive meals in women with

normal glucose metabolism (NGM) and with type 2 diabetes (DM2).

Methods: Post-menopausal women, 76 with NGM and 41 with DM2, received two

consecutive fat-rich meals and carbohydrate-rich meals on separate occasions.

Blood samples were taken before and at t=1, 2, 4, 6 and 8 hours following

breakfast; lunch was given at t=4h.

Results: In NGM women, fasting triglycerides, HbA1c, total cholesterol, and,

inversely, HDL-cholesterol were independently associated with TG-iAUC, and age

and fasting triglycerides were independently associated with glucose-iAUC (GL-

iAUC). In women with DM2, fasting triglycerides were independently associated

with TG-iAUC while HbA1c and fasting glucose were stronger than fasting

triglycerides associated with GL-iAUC. GL-iAUC and TG-iAUC were associated with

each other in women with DM2, but not in NGM.

Conclusions: In conclusion, GL-iAUC and TG-iAUC were correlated with each other

in women with DM2 and fasting triglycerides were associated with GL-iAUC and TG-

iAUC. These findings suggest a common underlying mechanism for postprandial

increments in glucose and triglycerides, especially occurring in women with DM2.

In healthy women, elevated TG-iAUC may be part of a dyslipidemic lipid profile,

whereas the relation between fasting triglycerides and GL-iAUC is suggestive for an

association with insulin resistance.

POSTPRANDIAL GLUCOSE AND TRIGLYCERIDES

29

INTRODUCTION

Already in 1979, Zilversmit postulated that atherogenesis might be a postprandial

phenomenon [1]. Since then, both postprandial hyperglycemia and

hypertriglyceridemia have been identified as risk markers for cardiovascular

disease (CVD) [2,3]. Glucose values two hours following an oral glucose tolerance

test (OGTT) have been shown to be better indicators of CVD risk than fasting

glucose concentration in the general population [4]. We recently demonstrated that

postprandial glucose levels were stronger than the fasting levels, associated with

carotid intima-media thickness in women with normal glucose metabolism (NGM)

[5]. Others have shown that postprandial hypertriglyceridemia is more strongly

related to carotid intima-media thickness than the fasting triglyceride levels [6,7].

The determinants of these postprandial glucose and triglyceride responses are not

well known but might, at least in part, overlap. However, the possible determinants

of postprandial triglyceride and glucose concentrations have not been described in

a single study population before. To date, most studies used a single, often

artificially composed liquid fat and/or carbohydrate load to assess postprandial

responses. Since a first meal can affect the glucose and triglyceride response to a

second meal, we chose to apply two consecutive meals [8,9]. Postmenopausal

women were invited for the present study because in these women, postprandial

triglyceride responses have been shown to be elevated [10], and type 2 diabetes

(DM2) was shown to confer a higher relative risk for CVD as compared to men [11].

In light of the above considerations, we assessed associations of clinical and

biochemical variables with postprandial triglyceride and glucose day profiles in

postmenopausal women with NGM and DM2, in order to investigate whether

postprandial hypertriglyceridemia and hyperglycemia are the result of a similar

underlying mechanism.

SUBJECTS AND METHODS

Study population

The study population has been described in detail previously [5]. In brief, women

with DM2 were randomly selected from the registry of the Diabetes Care System in

the city of Hoorn, the Netherlands. Women with NGM were randomly selected from

the municipal registry of the city of Hoorn. All women were between 50 and 65

years of age, were post-menopausal, non-smokers, had no untreated endocrine

CHAPTER 2

30

disorder other than DM2, did not use HMG-CoA reductase inhibitors, short acting

insulin analogues, peroxisome proliferator-activated receptor-α and/or γ agonists,

oral corticosteroids, or hormone replacement therapy. Women without known DM2

underwent an OGTT and were selected on NGM status (fasting glucose <6.1, 2-hour

post-load glucose <7.8 mmol/L [13]). Women with DM2 were excluded if they had

HbA1c >9.0%. Of the 1,063 women who were invited, a total of 431 of the women

were total non-responders and 258 women were unwilling to participate. A total of

257 women did not meet the inclusion criteria.

Finally, 76 women with NGM and 41 women with DM2 completed the study

protocol. All women gave written informed consent. The study was approved by the

ethics committee of the VU University Medical Center, Amsterdam, the Netherlands.

Measurements

The study consisted of a screening visit and two separate visits for the test-meals

with a minimum interval of one week and a maximum interval of one month

between visits. On the screening visit, blood samples were drawn after a 12-h

overnight fast to determine levels of HbA1c, plasma glucose, total cholesterol,

triglycerides, alanine aminotransferase (ALT) and creatinine. Women who were

selected from the municipal registry underwent an OGTT.

Blood pressure was measured at the left upper arm three times with 5 minutes

intervals using an oscillometric blood pressure measuring device (Collin Press-mate

BP-8800, Colin, Komaki-City, Japan) after a 15-minute supine rest. Weight and

height were measured twice in barefooted participants wearing light clothes only.

BMI was calculated as weight (in kg) divided by the square of height (in m). Waist

circumference was measured twice at the level midway between the lowest rib

margin and the iliac crest, and hip circumference was measured at the widest level

over the greater trochanters. Medical history, medication, (former) smoking and

alcohol use were assessed by a questionnaire [14]. Finally, habitual physical activity

was assessed by the Short Questionnaire to Assess Health-enhancing physical

activity of which reproducibility and relative validity were described previously [15].

On the second and the third visit, women arrived at the test facility in the morning

after an overnight fast. They had abstained from exercise 24 hours before the

study visit. Blood samples were taken before (twice) and at t=1, t=2, t=4, t=6 and

t=8 hours after ingestion of the first test meal. The second test meal was given at

t=4 hour, immediately after the blood sample was taken.


31

Test meal composition

Postprandial meal responses were examined following the consumption of two

standardized test meals (breakfast and lunch) on two separate occasions, either

with a high fat content or a high carbohydrate content. The fat-rich and the

carbohydrate-rich occasions were performed in random order. Each portion of the

ingredients was weighed before the meal was prepared. The nutrient composition

of the meals was calculated from the Dutch Food Composition Tables [16]. The fat-

rich meals (both breakfast and lunch) consisted of 2 croissants, 10 g of butter, 40

g of fat-rich cheese and 300 ml of fat-rich milk (3349 kJ; 50 g fat; 56 g

carbohydrates and 28 g proteins). The carbohydrate-rich meals (both breakfast and

lunch) consisted of 2 slices of bread, 25 g of marmalade, 30 g of cooked chicken

breast, 50 g of ginger bread and 300 ml of drinkable yogurt enriched with 45 g of

soluble sugars (3261 kJ; 4 g fat, 162 g carbohydrates and 22 g of proteins). Both

meals were eaten within 10 minutes. Apart from the test meals and water (ad

libitum), participants refrained from food and drinks. In addition, physical activity

was limited during the day.

Laboratory analysis

All laboratory analyses were performed at the VU University Medical Center

(department of Clinical Chemistry) in Amsterdam, the Netherlands. Serum total

cholesterol, HDL-cholesterol and triglycerides were measured by enzymatic

colorimetric assays (Roche, Mannheim, Germany). LDL-cholesterol was calculated

according to the Friedewald-formula [17]. ALT was determined by an enzymatic

assay (Roche, Mannheim, Germany) according to the methods proposed by the

International Federation of Clinical Chemistry and Laboratory Medicine [18]. Plasma

glucose concentration was determined with a glucose oxidase method (Granutest,

Merck, Darmstadt, Germany) and HbA1c was measured with cation-exchange

chromatography (Menarini Diagnostics, Florance, Italy). The inter-assay coefficients

of variation for the triglyceride and glucose measurements were <1.8% and <2.2%,

respectively. Immunospecific insulin was measured in serum by an immunometric

assay in which proinsulin does not cross-react (ACS Centaur, Bayer Diagnostics,

Mijdrecht The Netherlands). The inter- and intra-assay coefficients of variation for

insulin were 6% and 3%, respectively.

CHAPTER 2

32

Statistical analyses

Analyses were performed by SPSS for Windows 10 (SPSS Inc. Chicago, IL). Data are

presented as mean values (standard deviation), percentage or, in case of skewed

distribution, as median values (interquartile range) unless otherwise indicated.

Differences in characteristics between NGM and DM2 were tested with t-test for

continuous variables and with χ2-test for dichotomous variables. Postprandial

responses were calculated as incremental area under the curve (iAUC) with the

trapezoid method. Missing values for a given time-point (0.5% of all the glucose

and triglyceride values) were imputed by interpolation. Insulin resistance was

estimated by homeostasis model assessment (HOMA-IR), calculated as (mean

fasting insulin(mU/mL)*mean fasting glucose(mmol/L))/22.5 [19]. Mean glucose

and insulin concentrations were derived from fasting measurements on two

separate study days. To study the associations of possible determinants of

postprandial triglyceride and glucose responses, linear regression analysis was

performed with adjustment for age. As dependent variables, postprandial

triglycerides (iAUC) after the fat-rich meals and postprandial glucose concentrations

(iAUC) after the carbohydrate-rich meals were used. The associations were

expressed as standardized regression coefficients (95% CI) by dividing the

dependent and independent variables by the SD derived from the entire study

population. A regression coefficient of 0.5 means that if the independent variable

increases by 1.0 SD, the dependent variable increases by 0.5 SD. Multivariate

models included age and all variables that were associated with the outcome

variables at level P < 0.10. For all other analysis, we considered a two-sided P-value

<0.05 to indicate statistical significance.

RESULTS

Characteristics of the study population

Clinical and biochemical characteristics of the participants are listed in Table 1.

Figure 1 shows the 8h-time courses of triglyceride and glucose concentrations

following two consecutive fat-rich and two consecutive carbohydrate-rich meals.

Triglyceride-iAUC (TG-iAUC) after the fat-rich meals was similar in the two groups

(P=0.33) (Table2). Following the carbohydrate-rich meals, TG-iAUC was most

marked in women with DM2 compared to NGM women (P=0.01). As expected,

glucose-iAUC (GL-iAUC) and insulin-iAUC after the fat-rich and the carbohydrate-rich

meals was most marked in the women with DM2 (All P<0.01).


33

Table 2.1. Clinical and biochemical characteristics of the 117 participants a

NGM DM2

N 76 41

Age (years) 60.1 (4.0) 58.9 (3.7)

Duration of DM2 (years) NA 5 (3-9)

Anthropometry

BMI (kg/m2) 26.3 (3.6) 32.7 (6.0) b

Waist (m) 0.88 (0.10) 1.04 (0.14) b

Hip (m) 1.04 (0.08) 1.12 (0.12) b

Glucose metabolism

HbA1c (%) 5.6 (0.3) 6.6 (0.6) b

Fasting insulin (pmol/L) c 33.2 (25.5-47.5) 82.7 (38.8-122.5) b

HOMA-IR c 1.29 (0.94-1.95) 4.05 (1.92-7.07) b

Lipids (mmol/L)

Total cholesterol 6.0 (0.9) 5.6 (1.0)

HDL-cholesterol 1.80 (0.49) 1.51 (0.34) b

LDL-cholesterol 3.7 (0.9) 3.2 (1.0) b

Blood pressure (mmHg)

Systolic 131 (15) 143 (16) b

Diastolic 72 (8) 79 (7) b

Liver enzyme (U/L)

Alanine aminotransferase 20 (16-26) 27 (20-40) b

Medication (%)

Antihypertensive medication 16 66 b

Blood glucose lowering medication NA 71

Use of insulin NA 17

Lifestyle

Former smoking (%) 49 39

Habitual physical activity (h/week) 35 (23-46) 35 (22-46)

Alcohol > 0 gram/day (%) 78 37 b

NGM: Normal Glucose Metabolism, DM2: Diabetes Mellitus, HOMA-IR: Insulin resistance estimated

by Homeostasis Model Assessment. Differences between groups tested with t-test for continuous

variables and with χ2-test for dichotomous variables. a Data presented as mean values (SD) or

percentages. In case of skewed distribution, data presented as median (interquartile range) and Ln-

transformed values were tested. b P < 0.05. c Subjects using insulin were excluded and fasting

insulin was calculated as mean of two measurements.

CHAPTER 2

34

Fat-rich meals

0 2 4 6 8

Triglycerides (mmol/l)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

B L

Carbohydrate-rich meals

0 2 4 6 8

Triglycerides (mmol/l)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

B L

0 2 4 6 8

Glucose (mmol/l)

0

2

4

6

8

10

12

14

16

B L

0 2 4 6 8

Glucose (mmol/l)

0

2

4

6

8

10

12

14

16

B L

Time (hours)

0 2 4 6 8

Insulin (pmol/l)

0

200

400

600

800

1000

1200 B L

Time (hours)

0 2 4 6 8

Insulin (pmol/l)

0

200

400

600

800

1000

1200B L

Figure 3.1. Triglyceride, glucose and insulin concentrations (mean±SEM) after ingestion of

two consecutive meals. For insulin figures, patients who use insulin (n=7) were excluded.

B: breakfast, L: lunch. NGM: normal glucose metabolism (○), DM2: type 2 diabetes mellitus

(●).


35

Table 2.2. Fasting and postprandial glucose and triglyceride levels

after fat-rich meals and after carbohydrate-rich meals a

NGM DM2

Fat-rich meals

Fasting triglycerides

(mmol/L)

1.2 (0.6) 1.7 (0.7) b

TG-iAUC (mmol/L) 0.9 (0.5) 1.0 (0.5)

Fasting glucose (mmol/L) 5.3 (0.4) 7.2 (1.5) b

GL-iAUC (mmol/L) 0.1 (0.5) 0.7 (1.5) b


Fasting triglycerides

(mmol/L)

1.2 (0.6) 1.7 (0.6) b

TG-iAUC (mmol/L) 0.2 (0.3) 0.3 (0.3) b

Fasting glucose (mmol/L) 5.2 (0.4) 7.1 (1.3) b

GL-iAUC (mmol/L) 0.4 (0.8) 3.6 (2.9) b

NGM: Normal Glucose Metabolism, DM2: Diabetes Mellitus, TG: triglycerides,

iAUC: incremental area under the curve, GL: glucose. TG-iAUC and GL-iAUC

were presented as mean iAUC during the day. a Data presented as mean

values (SD). b T-test P < 0.05.

Associations with postprandial triglycerides

Results of linear regression analyses are presented in Table 3. In the NGM group,

fasting triglycerides were associated with TG-iAUC; also, HbA1c, fasting insulin,

HOMA-IR, total cholesterol were positively associated, whereas HDL-cholesterol and

hip circumference were negatively associated with TG-iAUC. To study whether

potential determinants of TG-iAUC were independent of fasting triglyceride levels,

we made a multivariate model. For women with NGM, we included fasting

triglycerides, hip circumference, HbA1c, fasting insulin, total cholesterol and HDL-

cholesterol and adjusted for age and BMI (R2=0.52), and all these variables, except

fasting insulin (P=0.08) and hip circumference (P=0.07), remained statistically

significantly associated with TG-iAUC. In the DM2 group, fasting triglyceride, HbA1c

and age were the strongest predictors of TG-iAUC. In a multivariate model with

fasting triglycerides, HbA1c and age (R2=0.29) fasting triglycerides was the only

independent determinant of TG-iAUC.

CHAPTER 2

36

Table 2.3. Linear regression analysis of clinical and biochemical parameters with postprandial triglyceride and glucose concentrations

NGM (n=76) DM2 (n=41)

Variables (SD) a

TG-iAUC GL-iAUC TG-iAUC GL-iAUC

Age (3.9 years) 0.04 (-0.18;0.25) 0.07 (-0.003;0.14) b,c 0.40 (0.04;0.75) b -0.04 (-0.46;0.39)

DM2 duration (4.3 years) - - 0.02 (-0.32;0.35) 0.18 (-0.22;0.58)

BMI (5.4 kg m-2) 0.05 (-0.29;0.39) 0.04 (-0.07;0.16) 0.07 (-0.24;0.37) 0.27 (-0.09;0.62)

Waist (0.14 m) d

0.41 (-0.15;0.97) 0.20 (0.01;0.38)b

0.34 (-0.62;1.29) 0.22 (-0.90;1.34)

Hip (0.10 m) d

-0.75 (-1.21;-0.29)b -0.13 (-0.29;0.03) 0.43 (-0.32;1.18) 0.06 (-0.82;0.94)

Fasting glucose (1.2 mmol/L) 0.50 (-0.21;1.21) -0.08 (-0.32;0.16) 0.11 (-0.20;0.41) 0.30 (-0.05;0.65)b

HbA1c (0.7%) 0.77 (0.31;1.22)b,c 0.13 (-0.03;0.29) 0.33 (-0.01;0.67)

b0.86 (0.54;1.18)

b,c

Fasting insulin (ln) (0.67) e 0.39 (0.05;0.74) b 0 (-0.12;0.12) 0.23 (-0.23;0.68) 0.47 (0.001;0.93)

b

HOMA-IR (ln) (0.79) e 0.32 (0.05;0.59) b 0 (-0.09;0.10) 0.18 (-0.16;0.51) 0.42 (0.09;0.74)

b

Total cholesterol (1.0 mmol/L) 0.32 (0.09;0.55) b,c -0.01 (-0.09;0.07) 0.24 (-0.06;0.54) 0.33 (-0.03;0.68)b

HDL-cholesterol (0.47 mmol/L) -0.31 (-0.51;-0.12) b,c -0.10 (-0.17;-0.04) b 0.01 (-0.47;0.50) -0.16 (-0.74;0.42)

Fasting triglycerides (0.6 mmol/L) 0.60 (0.39;0.81)b,c 0.13 (0.05;0.21)

b,c0.40 (0.06;0.73)

b,c0.48 (0.08;0.88) b

ALT (ln) (0.45) 0.09 (-0.16;0.33) 0.01 (-0.08;0.09) 0.03 (-0.30;0.36) 0.31 (-0.07;0.70)

Systolic BP (16 mmHg) f

0.14 (-0.13;0.41) 0.04 (-0.05;0.13) 0.17 (-0.18;0.52) 0.07 (-0.37;0.51)

Alcohol (65 g/week) 0.06 (-0.15;0.27) 0.02 (-0.05;0.09) 0.11 (-0.40;0.62) -0.47 (-1.06;0.12)

Habitual physical activity (18 h/week) 0 (-0.23;0.23) -0.05 (-0.12;0.03) -0.33 (-0.72;0.07) -0.42 (-0.90;0.06)b

NGM: Normal Glucose Metabolism, DM2: Diabetes Mellitus, TG: triglycerides, iAUC: incremental area under the curve, HOMA -IR: Insulin resistance estimated by Homeostasis

Model Assessment, ALT: alanine aminotransferase.

a Standardized regression coefficients (95% CI) adjusted for age in SD per 1 SD increase in independent variable. SD for TG -iAUC 0.5 mmol/L and for GL -iAUC 2.3 mmol/L. b

P < 0.10, variable included in the multivariate model; insulin and not HOMA- -IR were included when both were significant. c variable significant at level P < 0.05 in

multivariate model. dAdditionally adjusted for BMI.

eSubjects using insulin (n=7) were excluded.

fAdditionally adjusted for use of antihypertensive medication.


NGM (n=76) DM2 (n=41)

Variables (SD) a


Age (3.9 years) 0.04 (-0.18;0.25) 0.07 (-0.003;0.14) b,c 0.40 (0.04;0.75) b -0.04 (-0.46;0.39)


BMI (5.4 kg m-2) 0.05 (-0.29;0.39) 0.04 (-0.07;0.16) 0.07 (-0.24;0.37) 0.27 (-0.09;0.62)

Waist (0.14 m) d

0.41 (-0.15;0.97) 0.20 (0.01;0.38)b


NGM (n=76) DM2 (n=41)

Variables (SD) a


Age (3.9 years) 0.04 (-0.18;0.25) 0.07 (-0.003;0.14) b,c 0.40 (0.04;0.75) b -0.04 (-0.46;0.39)


BMI (5.4 kg m-2) 0.05 (-0.29;0.39) 0.04 (-0.07;0.16) 0.07 (-0.24;0.37) 0.27 (-0.09;0.62)

Waist (0.14 m) d

0.41 (-0.15;0.97) 0.20 (0.01;0.38)b

0.34 (-0.62;1.29) 0.22 (-0.90;1.34)

Hip (0.10 m) d

-0.75 (-1.21;-0.29)b -0.13 (-0.29;0.03) 0.43 (-0.32;1.18) 0.06 (-0.82;0.94)


HbA1c (0.7%) 0.77 (0.31;1.22)b,c 0.13 (-0.03;0.29) 0.33 (-0.01;0.67)

b0.86 (0.54;1.18)

b,c


b

HOMA-IR (ln) (0.79) e 0.32 (0.05;0.59) b 0 (-0.09;0.10) 0.18 (-0.16;0.51) 0.42 (0.09;0.74)

b




b,c0.40 (0.06;0.73)

b,c0.48 (0.08;0.88)

0.34 (-0.62;1.29) 0.22 (-0.90;1.34)

Hip (0.10 m) d

-0.75 (-1.21;-0.29)b -0.13 (-0.29;0.03) 0.43 (-0.32;1.18) 0.06 (-0.82;0.94)


HbA1c (0.7%) 0.77 (0.31;1.22)b,c 0.13 (-0.03;0.29) 0.33 (-0.01;0.67)

b0.86 (0.54;1.18)

b,c


b

HOMA-IR (ln) (0.79) e 0.32 (0.05;0.59) b 0 (-0.09;0.10) 0.18 (-0.16;0.51) 0.42 (0.09;0.74)

b




b,c0.40 (0.06;0.73)

b,c0.48 (0.08;0.88) b

ALT (ln) (0.45) 0.09 (-0.16;0.33) 0.01 (-0.08;0.09) 0.03 (-0.30;0.36) 0.31 (-0.07;0.70)

Systolic BP (16 mmHg) f

0.14 (-0.13;0.41) 0.04 (-0.05;0.13) 0.17 (-0.18;0.52) 0.07 (-0.37;0.51)

Alcohol (65 g/week) 0.06 (-0.15;0.27) 0.02 (-0.05;0.09) 0.11 (-0.40;0.62) -0.47 (-1.06;0.12)

Habitual physical activity (18 h/week) 0 (-0.23;0.23) -0.05 (-0.12;0.03) -0.33 (-0.72;0.07) -0.42 (-0.90;0.06)b

NGM: Normal Glucose Metabolism, DM2: Diabetes Mellitus, TG: triglycerides, iAUC: incremental area under the curve, HOMA -IR: Insulin resistance estimated by Homeostasis

Model Assessment, ALT: alanine aminotransferase.

a Standardized regression coefficients (95% CI) adjusted for age in SD per 1 SD increase in independent variable. SD for TG -iAUC 0.5 mmol/L and for GL -iAUC 2.3 mmol/L. b

P < 0.10, variable included in the multivariate model; insulin and not HOMA- -IR were included when both were significant. c variable significant at level P < 0.05 in

multivariate model. dAdditionally adjusted for BMI.

eSubjects using insulin (n=7) were excluded.

fAdditionally adjusted for use of antihypertensive medication.


37

Associations with postprandial glucose

Results of the linear regression analyses for fasting and postprandial glucose after

the carbohydrate-rich meals are presented in Table 3. In the NGM group, age, waist

circumference and fasting triglycerides were positively, and HDL-cholesterol was

inversely associated with GL-iAUC. In a multivariate model containing age, waist

circumference, BMI, HDL-cholesterol and fasting triglycerides (R2=0.24), fasting

triglycerides were still associated with GL-iAUC and age became significantly

associated with GL-iAUC (beta=0.07 [0.01;0.14]). In women with DM2, HbA1c,

fasting insulin, HOMA-IR and fasting triglycerides were positively associated with

GL-iAUC. For this reason, fasting insulin was part of the multivariate model, and we

excluded women with DM2 who used insulin (n=7). In this model, only HbA1c

remained associated with GL-iAUC when adjusted for age, fasting glucose, fasting

insulin, total cholesterol, triglycerides and habitual physical activity in a

multivariate model (R2=0.68). Because HbA1c is not considered as a determinant of

GL-iAUC, we also made a multivariate model including the above mentioned

variables except HbA1c. In this model (R2=0.51), fasting glucose was the strongest

determinant of GL-iAUC (beta=0.49 [0.19;0.82]).

Association between postprandial triglyceride and glucose responses

We additionally assessed the age-adjusted association of GL-iAUC with TG-iAUC. In

NGM women, no statistically significant association was found between these

metabolic responses (beta=0.39 [-0.31;1.08]). In contrast, for women with DM2, GL-

iAUC and TG-iAUC were associated (beta=0.26 [0.002;0.53]). The latter association

was in part dependent of fasting triglycerides; the regression coefficient reduced to

0.18 [-0.10;0.45] when fasting triglycerides were added to the model with age and

GL-iAUC.

DISCUSSION

The present study is to our knowledge the first to assess both postprandial

triglyceride and glucose responses at two separate occasions in one study

population. We demonstrated that fasting triglycerides were associated with both

TG-iAUC and GL-iAUC, but other potential determinants of TG-iAUC and GL-iAUC

differed. In spite of similar TG-iAUC responses in NGM and DM2, determinants of

TG-iAUC and also of GL-iAUC clearly differed between women with NGM and DM2.

CHAPTER 2

38

We expected a more markedly prolonged triglyceride response especially after the

second meal in patients with DM2 [20]. The relative lack of exaggerated triglyceride

response in women with DM2 might be the result of the meal-composition. The

substantial amount of carbohydrates in the fat-rich mixed meal might have resulted

in a high insulin response. Indeed, attenuation of the triglyceride response was

previously shown when glucose was added to a liquid fat-load [21]. Also, the

patients with DM2 included in the present study were well-controlled, possibly

contributing to an ameliorated response of the liver to the meal-induced insulin

response. On the other hand, since the women with DM2 were more insulin

resistant than women with NGM (HOMA-IR), an elevated TG-iAUC was expected [22].

Finally, we cannot exclude the possibility that a prolonged triglyceride response in

patients with DM2 might have become evident with a longer observation period

[23].

A common mechanism for elevated postprandial triglyceride and glucose

responses?

Insulin resistance or central obesity, both components of the metabolic syndrome,

might contribute to increased postprandial triglyceride and glucose levels.

Resistance to the suppressive effect of insulin on hepatic VLDL-production results

in exaggerated postprandial triglyceride responses [24,25]. We found that fasting

insulin and HOMA-IR, as a reflection of insulin resistance and in particular hepatic

insulin resistance, were associated with TG-iAUC in women with NGM, but the

association disappeared when fasting triglycerides were added to the model.

Furthermore, no association of HOMA-IR and fasting insulin with TG-iAUC in women

with DM2 was found. However, also fasting triglycerides are a recognized

component of the metabolic syndrome [26], and the association between fasting

triglycerides and TG-iAUC might, at least in part, be the result of underlying hepatic

insulin resistance.

Impaired glucose tolerance has been attributed to a combination of peripheral

insulin resistance and beta-cell dysfunction, whereas impaired fasting glucose is

mainly associated with hepatic insulin resistance and impaired insulin secretion

[27]. This corroborates with our finding that fasting glucose was associated with

GL-iAUC in women with DM2, but not in NGM. The combination of hepatic insulin

resistance and beta-cell dysfunction, reflected by elevated fasting and postprandial

glucose levels, is common in DM2, but not in NGM.


39

Central obesity as a component of the metabolic syndrome may be the common

mechanism for elevated postprandial triglyceride and glucose concentration. The

increased flux of free fatty acids in central obesity and elevated levels of pro-

inflammatory adipocytokines result in so-called ectopic fat depositions in liver,

muscle and beta-cells contributing to the development of hepatic and peripheral

insulin resistance, and beta-cell dysfunction [28]. Studies in men reported an

association between visceral obesity and TG-iAUC [29,30]. In line with these

observations, a study among obese women showed that abdominal obesity, but not

obesity as such, was associated with TG-iAUC. In the present study, the association

between waist circumference and TG-iAUC was, although not statistically

significant, stronger than the association between BMI and TG-iAUC. Furthermore,

we found that hip circumference was inversely associated with TG-iAUC in women

with NGM. An inverse relationship between larger hip circumference and fasting

triglycerides was previously found in both men and women [31,32], and may be

explained by an enhanced buffering capacity for triglyceride storage which

prevents from fat deposition in other organs [33]. The findings for GL-iAUC are

comparable to those for TG-iAUC; central obesity seemed to be more strongly

associated with elevated postprandial glucose than BMI in NGM women.

The present data also suggest that an elevated postprandial triglyceride response

might be part of a dyslipidemic lipid profile in women with NGM. Independent of

fasting triglycerides, total cholesterol was positively and fasting HDL-cholesterol

concentration was inversely associated with TG-iAUC. It is known that high levels of

triglycerides and low HDL-cholesterol concentrations are associated [34] and

postprandial lipemia might be another feature of this so-called ‘diabetic’

dyslipidemia. An important observation is that TG-iAUC was associated with GL-

iAUC in DM2, but not in NGM. This suggests that the underlying mechanisms, e.g.

hepatic insulin resistance and/or beta-cell dysfunction, is present in DM2, but not

(yet) in NGM. Other potential mechanisms for postprandial triglyceride and glucose

responses. An interesting finding was that HbA1c was strongly associated with TG-

iAUC in women with NGM. It is known that improvement of glycemic control also

improves postprandial triglyceride concentrations in patients with DM2 [35]. This

effect is attributed to the lipoprotein lipase (LPL) mediated clearance of TG-rich

lipoproteins, which is hampered when insulin action and/or secretion is inadequate

[36]. Since we did not measure LPL activity in this study, we cannot verify this

assumption. The proportion of variance explained by the multivariate model for TG-

iAUC is lower for patients with DM2 (R2 = 0.29) compared to women with NGM (R2

= 0.52). In DM2, insulin resistance, only partly reflected by fasting triglycerides

CHAPTER 2

40

possibly plays an important role in increasing postprandial triglycerides, including

the earlier discussed decreased LPL activity. The proportion of variance explained

(R2) by the multivariate model for GL-iAUC was 0.51 for DM2 and 0.24 for the NGM

group. Obviously, the range in GL-iAUC is much smaller in women with NGM.

Furthermore, we did not consider the effect of gut hormones (incretins), rate of

gastric emptying and peripheral insulin resistance on GL-iAUC, which play

important roles in glucose regulation.

Study limitations

A number of potential limitations of this study should be considered. First, the

population consisted of Caucasian post-menopausal women aged 50 to 65, and

caution should be exercised to generalise our findings to other populations.

Second, the cross-sectional design limits us to assess causal relationships. Third,

as a result of the smaller sample size of the DM2 population as compared to the

NGM study population, less associations in the DM2 study population might have

reached statistical significance.

Conclusion

In conclusion, GL-iAUC and TG-iAUC were interrelated in women with DM2 and

fasting triglycerides were associated with both GL-iAUC and TG-iAUC. These

findings suggest, at least partly, a common underlying mechanism for postprandial

increments in glucose and triglycerides, especially occuring in women with DM2. In

healthy women, elevated TG-iAUC may belong to a dyslipidemic lipid profile

whereas fasting triglycerides were related to GL-iAUC, suggestive for an association

with insulin resistance.

ACKNOWLEDGEMENTS

The authors thank Jannet Entius, Tanja Nansink, Lida Ooteman and Marianne

Veeken for their excellent technical assistance, Jolanda Bosman for her support in

organizing the study at the Diabetes Research Center, and all participants for their

contributions. This research was financially supported by a grant from the Dutch

Diabetes Research Foundation (grant no.2001.00.052) and by funding from

Novartis International AG, Switzerland.


41

REFERENCES

1. Zilversmit DB. Atherogenesis: a postprandial phenomenon. Circulation 1979;60:473-

85.





4. de Vegt F, Dekker JM, Ruhe HG, et al. Hyperglycaemia is associated with all-cause and

cardiovascular mortality in the Hoorn population: the Hoorn Study. Diabetologia

1999;42:926-31.

5. Alssema M, Schindhelm RK, Dekker JM, et al. Postprandial glucose, and not

triglyceride concentrations, are associated with carotid intima media thickness in

women with normal glucose metabolism: The Hoorn prandial study. Atherosclerosis

2007 (in press).



2000;23:1401-6.

7. Sharrett AR, Chambless LE, Heiss G, et al. Association of postprandial triglyceride and

retinyl palmitate responses with asymptomatic carotid artery atherosclerosis in

middle-aged men and women. The Atherosclerosis Risk in Communities (ARIC) Study.


8. Silva KD, Wright JW, Williams CM, Lovegrove JA. Meal ingestion provokes entry of

lipoproteins containing fat from the previous meal: possible metabolic implications.

Eur J Nutr 2005;44:377-83.

9. Brighenti F, Benini L, Del Rio D, et al. Colonic fermentation of indigestible

carbohydrates contributes to the second-meal effect. Am J Clin Nutr 2006;83:817-22.

10. van Beek AP, Ruijter-Heijstek FC, Erkelens DW, de Bruin TW. Menopause is associated

with reduced protection from postprandial lipemia. Arterioscler Thromb Vasc Biol

1999;19:2737-41.




12. 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;15:539-53.

13. Mooy JM, Grootenhuis PA, de Vries H, et al. Prevalence and determinants of glucose

intolerance in a Dutch caucasian population. The Hoorn Study. Diabetes Care

1995;18:1270-3.

CHAPTER 2

42

14. Wendel-Vos GC, Schuit AJ, Saris WH, Kromhout D. Reproducibility and relative validity

of the short questionnaire to assess health-enhancing physical activity. J Clin

Epidemiol 2003;56:1163-9.

15. Voorlichtingsbureau voor de Voeding. NEVO Table Netherlands (Dutch Nutrient

Database). 2001.

16. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-

density lipoprotein cholesterol in plasma, without use of the preparative

ultracentrifuge. Clin Chem 1972;18:499-502.

17. Bergmeyer HU, Horder M, Rej R. International Federation of Clinical Chemistry (IFCC)

Scientific Committee, Analytical Section: approved recommendation (1985) on IFCC

methods for the measurement of catalytic concentration of enzymes. Part 3. IFCC

method for alanine aminotransferase (L-alanine: 2-oxoglutarate aminotransferase, EC

2.6.1.2). J Clin Chem Clin Biochem 1986;24:481-95.

18. Matthews DR, Hosker JP, Rudenski AS, et al. Homeostasis model assessment: insulin

resistance and beta-cell function from fasting plasma glucose and insulin

concentrations in man. Diabetologia 1985;28:412-9.



Eur J Nutr 2005;44:377-83.

20. Cohen JC, Berger GM. Effects of glucose ingestion on postprandial lipemia and

triglyceride clearance in humans. J Lipid Res 1990;31:597-602.

21. Lewis GF, Steiner G. Acute effects of insulin in the control of VLDL production in

humans. Implications for the insulin-resistant state. Diabetes Care 1996;19:390-3.

22. Tushuizen ME, Nieuwland R, Scheffer PG, et al. Two consecutive high-fat meals affect

endothelial-dependent vasodilation, oxidative stress and cellular microparticles in

healthy men. J Thromb Haemost 2006;4:1003-10.

23. Garg A. Insulin resistance in the pathogenesis of dyslipidemia. Diabetes Care

1996;19:387-9.

24. Lewis GF, Steiner G. Acute effects of insulin in the control of VLDL production in

humans. Implications for the insulin-resistant state. Diabetes Care 1996;19:390-3.

25. Reaven GM. Banting lecture 1988. Role of insulin resistance in human disease.

Diabetes 1988;37:1595-607.

26. Meyer C, Pimenta W, Woerle HJ, et al. Different mechanisms for impaired fasting

glucose and impaired postprandial glucose tolerance in humans. Diabetes Care

2006;29:1909-14.

27. Raz I, Eldor R, Cernea S, Shafrir E. Diabetes: insulin resistance and derangements in

lipid metabolism. Cure through intervention in fat transport and storage. Diabetes

Metab Res Rev 2005;21:3-14.

28. Wideman L, Kaminsky LA, Whaley MH. Postprandial lipemia in obese men with

abdominal fat patterning. J Sports Med Phys Fitness 1996;36:204-10.


43

29. Couillard C, Bergeron N, Prud'homme D, et al. Postprandial triglyceride response in

visceral obesity in men. Diabetes 1998;47:953-60.

30. Snijder MB, Zimmet PZ, Visser M, et al. Independent and opposite associations of

waist and hip circumferences with diabetes, hypertension and dyslipidemia: the

AusDiab Study. Int J Obes Relat Metab Disord 2004;28:402-9.

31. Snijder MB, Zimmet PZ, Visser M, et al. Independent association of hip circumference

with metabolic profile in different ethnic groups. Obes Res 2004;12:1370-4.

32. Frayn KN. Adipose tissue as a buffer for daily lipid flux. Diabetologia 2002;45:1201-

10.

33. Ginsberg HN, Zhang YL, Hernandez-Ono A. Regulation of plasma triglycerides in

insulin resistance and diabetes. Arch Med Res 2005;36:232-40.

35. Jeppesen J, Zhou MY, Chen YD, Reaven GM. Effect of metformin on postprandial

lipemia in patients with fairly to poorly controlled NIDDM. Diabetes Care

1994;17:1093-9.

36. Ginsberg HN. Efficacy and mechanisms of action of statins in the treatment of

diabetic dyslipidemia. J Clin Endocrinol Metab 2006;91:383-92.

CHAPTER 2

44

45

Chapter 3

POSTPRANDIAL GLUCOSE AND NOT TRIGLYCERIDE CONCENTRATIONS ARE ASSOCIATED WITH CAROTID INTIMA MEDIA THICKNESS IN WOMEN

WITH NORMAL GLUCOSE METABOLISM: THE HOORN PRANDIAL STUDY

Marjan Alssema1

Roger K. Schindhelm2


Michaela Diamant2

Piet J. Kostense1,3

Tom Teerlink4

Peter G. Scheffer4

Giel Nijpels1,5

Robert. J. Heine1,2

From the 1EMGO Institute and the Departments of 2Endocrinology/Diabetes Center,

3Clinical Epidemiology and Biostatistics, 4Clinical Chemistry, and 5General Practice,

VU University Medical Center, Amsterdam, The Netherlands.


Copyright © 2007 Elsevier Ireland Ltd.

Reprinted with permission from Elsevier.

CHAPTER 3

46

ABSTRACT

The present study aimed to compare the associations of postprandial glucose

(ppGL) and postprandial triglycerides (ppTG) with carotid intima media thickness

(cIMT) in women with normal glucose metabolism (NGM) and type 2 diabetes

(DM2). Post-menopausal women (76 with NGM, 78 with DM2), received two

consecutive fat-rich and two consecutive carbohydrate-rich meals on separate

occasions. Blood samples were taken before and 1, 2, 4, 6 and 8 hours following

breakfast; lunch was given at t=4. Ultrasound imaging of the carotid artery was

performed to measure cIMT. In women with NGM, an increase of 1.0 mmol/l

glucose following the fat-rich meals was associated with a 50 µm cIMT increase

(P=0.04), and following the carbohydrate meals, an increase of 1.8 mmol/l glucose

was associated with a 50 µm larger cIMT (P=0.08). These associations were not

explained by classical cardiovascular risk factors. However, no association between

ppGL and cIMT was found in women with DM2 and ppTG were not associated with

cIMT. The association between ppGL and cIMT in normoglycemic women suggests

that ppGL in the normal range is a marker or a risk factor for atherosclerosis.

Postprandial glucose levels might be a better indicator of risk than post-OGTT

glucose levels or triglyceride levels.

POSTPRANDIAL METABOLISM AND ATHEROSCLEROSIS

47

INTRODUCTION

Men with type 2 diabetes mellitus (DM2) have a two- to three-fold and women a

three- to four-fold increased risk of cardiovascular disease (CVD) [1,2]. The

increased risk for CVD in patients with DM2 is only partly explained by the classical

CVD risk factors hypertension, high LDL-cholesterol, low HDL-cholesterol and

smoking [3].

Disturbances in the postprandial metabolism are hypothesized to contribute to the

excess risk of CVD, because of the elevated concentrations of glucose, triglyceride-

enriched lipoproteins and inflammatory markers in the postprandial period [4].

Glucose levels after an oral glucose tolerance test (OGTT) have been shown to be

better indicators of CVD risk than fasting glucose levels in the general population

and in subjects with a family history of DM2 [5-7]. Also, postprandial triglyceride

levels were more strongly related to cIMT than fasting triglyceride levels in both

healthy subjects [8-10] and in patients with DM2 [11,12]. However, the relative

contributions of postprandial glucose (ppGL) levels and postprandial triglyceride

(ppTG) levels to atherosclerosis are not clarified yet.

Most postprandial studies applied a single fat-rich liquid-formula or an OGTT, but

the association of the more physiological meal-related glucose and triglyceride

responses with CVD might be more relevant. Furthermore, during daytime, glucose

excursions following a first meal are known to affect the responses after a second

meal [13]. A cumulative effect of the triglyceride excursions following breakfast

and lunch is likely to occur, especially in subjects with DM2 [14].

The aim of the present study was to investigate the relation of ppGL and ppTG

following two consecutive fat-rich meals or two consecutive carbohydrate-rich

meals on two separate days, to cIMT, and to investigate whether these associations

differ between women with normal glucose metabolism (NGM) and women with

DM2.

CHAPTER 3

48

METHODS

Study population

Women with DM2 (n=522), from the registry of the Diabetes Care System in the city

of Hoorn, the Netherlands, and women who were randomly selected from the

municipal registry of Hoorn (n=541) aged 50-65 years at the beginning of the

study, were invited to participate in the study. Exclusion criteria were non-post-

menopausal status (menses in the last 12 months), smoking, untreated endocrine

disorder other than DM2, use of short acting insulin analogues, use of PPAR-α and -

γ agonists, use of oral corticosteroids, use of hormone replacement therapy,

fasting cholesterol >8.0 mmol/l, fasting triglycerides >4.0 mmol/l, systolic blood

pressure>190 mmHg, liver or renal impairment (liver derived enzymes >2.5 time

the upper limit of the laboratory reference range, creatinine >120 µmol/l). Women

who were selected from the municipal registry underwent a 75-g OGTT to verify

their glucose tolerance status (fasting glucose ≤6.0 mmol/l and 2-hour post-load

glucose <7.8 mmol/l [15]). Women with DM2 were excluded if they had HbA1c

>9.0%. Women with NGM were excluded if they used HMG-CoA reductase inhibitors

(statins), and women with DM2 who used statins were considered as a separate

study group. Of the 1,063 women who were invited, 431 women were complete

non-responders and 258 women were not willing to participate. A total of 220

women did not meet the inclusion criteria. Reasons for exclusion were: smoking

n=65, elevated fasting and/or post-load glucose levels (NGM only) n=33, pre-

menopausal n=28, drop-out n=22, failure to draw blood samples from a canula

n=19, use of hormone replacement therapy n=13, use of short acting insulin

analogues n=12, illness during study participation n=6, elevated triglycerides n=4,

use of PPAR-γ agonists n=4, use of statins (NGM only) n=3, blood pressure n=3,

miscellaneous reasons n=4, cholesterol n=1, creatinine n=1, liver enzymes n=1 and

missing cIMT n=1. Finally, 76 women with NGM and 78 women with DM2

completed the study protocol. All women gave written informed consent. The study

was approved by the ethics committee of the VU University Medical Center.

Screening visit

The study consisted of three separate visits, i.e. a screening visit and two visits for

the test-meals, with a maximum of one month apart. On the screening visit, fasting

blood samples were drawn after a 12-h overnight fast.


49

Blood pressure was measured at the left arm three times with 5 minutes intervals

with an oscillometric blood pressure measuring device (Collin Press-mate BP-8800,

Colin, Komaki-City, Japan) after a 15-minute supine rest. Weight and height were

measured twice in barefooted participants wearing light clothes only. BMI was

calculated as weight (in kg) divided by the square of height (in m). Waist

circumference was measured twice at the level midway between the lowest rib

margin and the iliac crest, and hip circumference was measured at the widest level

over the greater trochanters. Medical history, medication, smoking and alcohol

consumption were assessed by a questionnaire [16]. Finally, physical activity was

assessed by the Short Questionnaire to Assess Health-enhancing physical activity

[17].

Ultrasound imaging

A single observer (MA) performed ultrasound imaging with an ultrasound scanner

(350 Series, Pie Medical, Maastricht, The Netherlands), equipped with a 7.5-MHz

linear probe. Measurements were performed in the right common carotid artery at

10 mm proximal to the carotid bulb. Images were registered and analyzed by a

personal computer equipped with vessel wall movement detection software and an

acquisition system (Wall Track System, Pie Medical, Maastricht, the Netherlands).

After a 15-minute supine rest, the artery was visualized in B-mode. After defining

the segment, 10 mm proximal to the carotid bulb, the screen was switched to the

M-mode and data acquisition in real-time presentation on the computer screen was

enabled. Data were obtained by 3 consecutive measurements of 4 seconds each,

triggered by the R-top of a simultaneously recorded ECG [18]. Carotid IMT was

measured from the posterior wall, as the distance from the first to the second

echogenic line [19]. Reproducibility of scanning was assessed in 7 women with

DM2 and in 5 women with NGM, who were examined twice, two weeks apart. The

intra-observer coefficient of variation (=SD of the mean difference / (√2 * pooled

mean) * 100%) for cIMT was 5.2%.

Test meals

All participants received two consecutive fat-rich meals on one occasion and two

consecutive carbohydrate-rich meals on another occasion, in random order. Women

arrived at the test facility in the morning after an overnight fast. Blood samples

were taken before and at t=1, t=2, t=4, t=6 and t=8 hour after ingestion of the first

test meal (breakfast). Lunch was given at t=4, immediately after a blood sample

CHAPTER 3

50

was taken. The nutrient composition of the meals was calculated from the Dutch

Nutrient Database [20]. The fat-rich meals consisted of croissants, butter, cheese

and fat-rich milk (3349 kJ; 50 g fat; 56 g carbohydrates and 28 g proteins). The

carbohydrate-rich meals consisted of bread, marmalade, cooked chicken breast,

ginger bread and drinkable yogurt enriched with soluble carbohydrates (3261 kJ; 4

g fat, 162 g carbohydrates and 22 g proteins). Meals were eaten within 10 minutes.

Apart from the test meals and water (ad libitum), participants refrained from food

and drinks. In addition, throughout the visit, physical activity was limited to a short

walk between two adjacent rooms.

Laboratory analysis

All laboratory analyses were performed at the VU University Medical Center in

Amsterdam, the Netherlands. Serum total cholesterol, HDL-cholesterol and

triglycerides were measured by enzymatic calorimetric assays (Roche, Mannheim,

Germany). The inter-assay coefficients of variation for the triglyceride levels were

2.2% at a mean level of 1.0 mmol/l and 1.8% at a mean level of 1.9 mmol/l. Fasting

LDL-cholesterol was calculated according to the Friedewald-formula [21]. Plasma

glucose levels were determined with a glucose oxidase method (Granutest, Merck,

Darmstadt, Germany) and HbA1c was measured with reversed-phase cation

exchange chromatography (Menarini Diagnostics, Florence, Italy). The inter-assay

coefficients of variation for the glucose levels were 1.2% at a mean level of 4.9

mmol/l and 1.3% at a mean level of 19.0 mmol/l. Specific insulin was measured in

serum by an immunometric assay in which proinsulin did not cross-react (Advia

Centaur, Bayer Diagnostics, Mijdrecht, The Netherlands). The intra- and inter-assay

coefficients of variation for insulin were <4% and <7%, respectively.


Analyses were performed in SPSS 12.0.1 for Windows (SPSS Inc. Chicago, IL).

Differences in baseline characteristics between the three groups (NGM, DM2 and

DM2-ST) were tested with ANOVA with post-hoc analyses (LSD) for continuous

variables that were ln-transformed in case of skewed distribution. Dichotomous

variables were tested with χ2-test. Postprandial responses were calculated with the

trapezoid method as total area under the curve (AUC) divided by 8 (hours). Insulin

resistance was estimated by homeostasis model assessment (HOMA-IR), calculated

as (mean fasting insulin (µU/ml) * mean fasting glucose (mmol/l))/22.5 [22]. Means

for glucose and insulin were derived from fasting measurements on two separate


51

study days. To study the possible associations between risk factors and cIMT and

between postprandial responses and cIMT, we performed linear regression analysis

with adjustment for age. The associations were expressed as regression

coefficients per 1 SD (of the total study population) increase in independent

variable. We tested for possible interactions of independent variables with DM2 and

use of statins by calculating the p values of the interaction terms. Subsequently,

regression coefficients for ppGL and ppTG were adjusted for fasting levels to

investigate whether fasting levels explained the possible association between

postprandial response and cIMT. To investigate the relative contribution of ppTG

and ppGL to cIMT, we also made a mutually adjusted model for ppTG and ppGL.

Finally, we tested BMI, total cholesterol, fasting insulin, HOMA-IR, systolic blood

pressure, smoking, LDL-cholesterol and HDL-cholesterol as potentially confounding

factors. We considered a two-sided P-value <0.05 to indicate statistical significance.

For interaction terms, we considered P<0.15 statistically significant.

RESULTS

Clinical and biochemical characteristics assessed at the screening visit and fasting

and postprandial glucose and triglycerides were presented separately for women

with NGM, DM2 and women with DM2 using statin medication (DM2-ST, Table 1).

Fasting and postprandial triglycerides and glucose, were similar in the DM2 group

and in the DM2-ST group, but were higher compared to the NGM group (all p<0.01,

Table 3.1). Carotid IMT was increased and CVD was more prevalent in women with

DM2 who were not using statins, but these differences did not reach statistical

significance.

In Table 3.2, results of linear regression analysis of known CVD risk factors with

regard to cIMT are shown. Because no interaction with DM2 status or use of statins

was found for most of the risk factors, we combined the subgroups for these

analyses. Age was associated with cIMT: the estimated beta was 20.2, indicating

that a 4 year difference in age is associated with a 20 µm larger cIMT (P=0.02). Also

systolic blood pressure was associated with cIMT, a difference of 16 mmHg in

systolic blood pressure was associated with a 19 µm difference in cIMT (P=0.04).

However, the latter association became non-significant after adjustment for DM2

status and use of statins.

CHAPTER 3

52

Table 3.1. Population characteristics (N=154)

NGM DM2 DM2-ST

N 76 41 37

Age (years) 60.1 (4.0) 58.9 (3.7) 61.1 (4.0) a

Duration of DM2 (years) - 5 (3-9) 4 (2-9)

BMI (kg/m2) 26.3 (3.6) 32.7 (6.0) b 30.6 (4.6) b

Waist (cm) 88 (10) 104 (14) b 102 (13) b

Hip (cm) 104 (8) 112 (12) b 108 (10) b

2-h post-load glucose (mmol/l) 5.4 (1.0) - -

HbA1c 5.6 (0.3) 6.6 (0.6) b 6.8 (0.7) b

Fasting insulin (pmol/l) c,d 33.2 (25.5-47.5) 82.7 (38.8-122.5) b 82.3 (53.3-107.3) b

HOMA-IR c,d 1.29 (0.94-1.95) 4.05 (1.92-7.07) b 4.33 (2.53-6.61) b

Total cholesterol (mmol/l) c 5.7 (0.8) 5.5 (1.0) 4.4 (0.8) a,b

HDL-cholesterol (mmol/l) c 1.62 (0.42) 1.36 (0.29) b 1.33 (0.29) b

LDL-cholesterol (mmol/l) c 3.5 (0.8) 3.4 (0.9) 2.3 (0.7) a,b

Systolic blood pressure (mmHg) 131 (15) 143 (16) b 134 (16) a

Former smoking (%) 49 39 43

Physical activity (hours/week) 37 (19) 36 (17) 31 (21) b

Alcohol intake >0gram/day (%) 78 37 b 46 b

Antihypertensive medication (%) 16 66 b 62 b

Diabetes treatment (%)

Diet alone - 24 14

Metformin only - 39 35

Sulfonylureas only - 7 14

Metformin and sulfonylureas - 10 19

Metformin and acarbose - 2 0

Use of insulin - 17 19

Meal measurements

Fasting triglycerides (mmol/l) c 1.2 (0.5) 1.7 (0.6) b 1.8 (0.6) b

Fasting glucose (mmol/l) c 5.2 (0.4) 7.1 (1.4) b 7.5 (1.2) b

Fat-rich meals:

PpTG (mmol/l) 2.1 (0.9) 2.7 (1.0) b 2.7 (0.9) b

PpGL (mmol/l) 5.4 (0.5) 7.9 (1.8) b 7.8 (1.6) b

Carbohydrate-rich meals:

PpTG (mmol/l) 1.4 (0.7) 2.0 (0.6) b 2.0 (0.7) b

PpGL (mmol/l) 5.7 (0.8) 10.7 (3.5) b 11.5 (3.6) b

Prevalent cardiovascular disease (%)e 22 34 26

Carotid Intima Media Thickness (µm) 764 (110) 790 (107) 769 (121)

Data as means (SD) or percentages. In case of skewed distribution, medians (interquartile range) are

presented and Ln-transformed values were tested. NGM= normal glucose metabolism, DM2= type 2 diabetes mellitus, DM2-ST= type 2 diabetes mellitus using statins, HOMA-IR=homeostasis model

assessment for insulin resistance, ppTG=postprandial triglycerides, ppGL=postprandial glucose. a

P<0.05 compared to DM2 group, b P<0.05 compared to NGM group, c mean of 2 measurements, d

women using insulin excluded, e prevalent cardiovascular disease defined as ankle-brachial-index

<0.90, ECG abnormalities (1.1-1.3, 4.1-4.3, 5.1-5.3 or 7-1), history of arterial surgery, cerebral

vascular event, angina, amputation, claudication or possible myocardial infarction.


53

Table 3.2. Linear regression analysis (beta (95% CI)) of risk factors and carotid Intima Media Thickness (n=154) Variable SD Model 1 Model 2

Age 4.0 years 20.2 (2.6;37.8) a 22.7 (4.8;40.6) a

DM2 (yes versus no) 17.2 (-17.9;52.3) 33.1 (-9.3;75.5)

DM2 duration b 4.7 years 12.7 (-11.1;36.5) 11.0 (-12.5;34.5)

BMI 5.4 kg/m2 11.4 (-6.2;29.0) 7.7 (-13.0;28.3)

Waist c 14.0 cm 13.0 (-31.6;57.6) 16.6 (-30.6;63.8)

Hip c 10.3 cm -8.3 (-45.4;28.8) -8.7 (-47.6;30.3)

2-hour post-load glucose d 1.0 mmol/l 2.1 (-22.8; 27.0) -

HbA1c 0.76 % 12.4 (-5.2;29.9) 16.1 (-10.2;42.3)

Fasting insulin b,e 59 pmol/l 1.7 (-19.2;22.6) -4.1 (-28.8;20.6)

HOMA-IR b,e 3.4 5.0 (-15.3;25.2) -0.1 (-26.8;26.7)

Total cholesterol 1.01 mmol/l 6.7 (-10.9;24.3) 7.4 (-13.6;28.3)

HDL-cholesterol 0.39 mmol/l -8.0 (-25.8;9.8) -5.6 (-24.6;13.5)

LDL-cholesterol 0.93 mmol/l 10.1 (-7.5;27.6) 12.4 (-9.1;33.9)

Systolic blood pressure f 16 mmHg 18.9 (0.6;37.1) a 17.1 (-1.8;35.9)

Former smoking g (yes versus no) 23.5 (-25.6;72.5) -

Alcohol intake b 63 gram/week 11.5 (-6.4;29.4) 17.7 (-1.1;36.6)

Physical activity 19 hours/week -12.3 (-31.7;7.0) -12.0 (-31.5;7.4)

Beta in µm intima media thickness per 1 SD increase in independent variable. DM2= type 2 diabetes

mellitus, HOMA-IR=homeostasis model assessment for insulin resistance. Model 1: Adjusted for age.

Model 2: Adjusted for age, DM2 and use of statins. a P<0.05, b ln-transformed for analyses, c waist and

hip are additionally adjusted for each other and for BMI, d after an OGTT in women with NGM only, e

women using insulin excluded, f additionally adjusted for use of antihypertensive medication, g

interaction with DM2 (P=0.08) and use of statins (P=0.03), betas for women with NGM only. Other betas

were: -50.6 (-117.4;16.1) and 59.2 (-22.4;140.7) for women with DM2 with DM2-ST, respectively.

Results of linear regression analysis of fasting and postprandial glucose and

triglycerides are shown in Table 3. Because we observed different associations of

ppGL with cIMT for women with DM2 and women with NGM after both the fat-rich

(P-value for interaction P=0.12) and the carbohydrate-rich meals (P-value for

interaction P=0.15), results are presented separately for women with NGM and

DM2. We also found an interaction between fasting glucose and use of statin

medication (P=0.07). Therefore, we performed the regression analysis of fasting

glucose with cIMT for women in the DM2-ST group separately.

CHAPTER 3

54

Table 3.3. Associations (beta (95% CI)) of fasting and postprandial glucose and

triglycerides with carotid Intima Media Thickness

NGM (n=76)

Variable (SD) Model 1 Model 1 +

fasting levels

Model 1 +

mutually adjusted a

Fasting

Triglycerides (0.6 mol/l) 2.3 (-26.0;30.7) - -0.8 (-29.9;28.2)

Glucose (1.4 mmol/l) 50.4 (-45.8;146.6) - 51.0 (-48.1;150.0)

Fat meals

PpTG (1.0 mmol/l) 3.9 (-23.1;30.9) 8.6 (-48.5;65.9) -3.7 (-31.0;23.6)

PpGL (1.8 mmol/l) 92.6 (6.1;179.1) 91.8 (-8.0;191.7) 95.7 (5.5;185.9)

Carbohydrate meals

PpTG (0.7 mmol/l) 10.0 (-16.7;36.6) 43.5 (-18.5;105.5) 1.3 (-27.3;29.9)

PpGL (3.7 mmol/l) 98.8 (-13.7;211.3) 89.9 (-35.9;215.7) 96.7 (-26.0;219.3)

DM2 (n=78)

Variable (SD) Model 1 Model 1 +

fasting levels

Model 1 +

mutually adjusted a

Fasting

Triglycerides (0.6 mol/l) -5.6 (-34.1;22.9) - -12.5 (-43.0;17.9)

Glucose (1.4 mmol/l) 38.6 (6.2;71.0) b - 51.6 (16.8;86.5) b

Fat meals

PpTG (1.0 mmol/l) -13.4 (-40.3;13.6) -26.8 (-73.3;19.7) -14.0 (-41.1;13.1)

PpGL (1.8 mmol/l) 8.6 (-18.5;35.6) c -4.6 (-40.1;30.9) 9.5 (-17.6;36.5)

Carbohydrate meals

PpTG (0.7 mmol/l) -10.1 (-39.8;19.6) -20.6 (-81.2;39.9) -12.7 (-42.7;17.2)

PpGL (3.7 mmol/l) 14.4 (-13.1;42.0) c 5.7 (-29.2;40.7) 16.2 (-11.7;44.2)

Beta in µm intima media thickness per 1 SD increase in independent variable. NGM= normal glucose

metabolism, DM2= type 2 diabetes mellitus, ppTG=postprandial triglycerides, ppGL=postprandial

glucose. Model 1: adjusted for age and (DM2 group) use of statins. a Mutually adjusted as follows:

Fasting triglycerides and fasting glucose; ppTG and ppGL following fat-rich meals; ppTG and ppGL

following carbohydrate-rich meals, b Interaction with use of statins (P=0.07) beta for women with

DM2, not using statins only. Beta for DM2-ST-group –13.6 (-62.8;35.5), c Interaction with DM2

P<0.15.


55

Fasting levels of glucose were statistically significantly associated with cIMT in

women with DM2. In women with NGM, this association was of similar magnitude,

although not statistically significant (Table 3). After mutual adjustment of fasting

glucose and fasting triglycerides, the association between fasting glucose and cIMT

in women with DM2 became stronger. After adjustment for CVD risk factors, (BMI,

total cholesterol, fasting insulin, HOMA-IR, systolic blood pressure, smoking, LDL-

cholesterol or HDL-cholesterol) the association between fasting glucose and cIMT

remained statistically significant (data not shown).

PpTG after both the test meals was not associated with cIMT in women with NGM

and not in women with DM2. However, PpGL after both the fat-rich meals (p=0.04)

and the carbohydrate-rich meals (borderline significant P=0.08) was associated with

cIMT in women with NGM. A 1.0 mmol/l higher glucose level during the day of the

fat-rich meals was associated with a 50 µm larger cIMT. In addition, after the

carbohydrate meals, a 1.8 mmol/l higher glucose level was associated with a 50 µm

larger cIMT. These associations only slightly changed after adjustment for fasting

glucose or ppTG. Also adjustment for BMI, total cholesterol, fasting insulin, HOMA-

IR, systolic blood pressure, smoking, LDL-cholesterol or HDL-cholesterol as

potentially confounding factors did not alter the associations between ppGL and

cIMT (data not shown). Age-adjusted IMT in quartiles of ppGL for women with NGM

is shown in Figure 3.1.

To further explore the association between ppGL and cIMT, we performed

regression analyses for each of the glucose measurements separately (1,2 and 4

hour) and for ppGL after breakfast only (AUC 0-4 hour) with regard to cIMT.

Although borderline significant, glucose levels at t=1 and t=2 hour after the fat-rich

meal were associated with cIMT (1.0 mmol/l glucose associated with a 19 µm larger

cIMT at t=1 and 32 µm at t=2, both P=0.05). PpGL, after one meal (AUC 0-4 hour)

was associated with cIMT similar to ppGL calculated after two consecutive meals. A

1.0 mmol/l higher glucose level after the fat-rich breakfast was associated with a

44 µm larger cIMT (P=0.03).

CHAPTER 3

56

Fat-rich meals

600

650

700

750

800

850

<5.0 5.0-5.3 5.3-5.7 >5.7

cIMT (µµ µµm)


600

650

700

750

800

850

<5.2 5.2-5.7 5.7-6.1 >6.1

cIMT (µµ µµm)

Figure 3.1. Age-adjusted cIMT (±SEM) in quartiles of mean postprandial glucose levels (mmol/l, 0-8 hour) following fat-rich meals (upper) and following carbohydrate-rich meals (lower) for women with NGM. P for trend derived from linear regression analyses with postprandial glucose levels as continuous variable: P<0.05 and P=0.08 for fat-rich and carbohydrate-rich meals, respectively.

DISCUSSION

In the present cross-sectional study, we compared the relative contribution of ppGL

and ppTG to cIMT as an early marker of atherosclerosis. We found that ppGL,

irrespective of the meal composition, was associated with cIMT in women with NGM


57

but not in women with DM2. The association between ppGL and cIMT remained

unaltered after adjustment for ppTG or other CVD risk factors.

Mild hyperglycemia has previously been discussed as an underestimated

cardiovascular risk factor [23]. However, in the present study, we also included

women with NGM, confirmed by OGTT. Remarkably, only in this normal range of

fasting and postprandial glucose levels, ppGL was associated with cIMT. In the

women with DM2, an association was found with fasting glycemia but not with

ppGL. Possibly, a high biological variation in ppGL has obscured the association

with cIMT in these patients with DM2. Another possibility is that treatment with

glucose lowering medication and dietary intervention in the women with DM2

might have influenced the association between ppGL levels and cIMT [24].

Based on previously published papers, we would have expected an association

between ppTG and cIMT, especially in patients with DM2 [11,12], but also in

healthy subjects [8-10]. Moreover, post-menopausal women have been shown to

have higher postprandial lipemia [25], which could potentially contribute to the

high risk of CVD in these women [26,27]. However, no associations between ppTG

and cIMT were found in the present study. As to ppTG, no differences were found

in a study with persons with and without CVD [28]. In contrast, in another study,

patients with coronary heart disease were found to have higher fasting triglycerides

and higher ppTG than subjects without coronary heart disease [29]. In the

Atherosclerosis Risk in Communities Study, ppTG levels were found to be

associated with early atherosclerosis, but only in non-obese subjects [8] and the

investigators suggested that this association might be explained by high levels of

atherogenic triglyceride-rich lipoprotein remnants in the postprandial state. These

lipoprotein remnants were indeed related to cIMT, independent of ppTG [30]. Since

we did not measure lipoprotein remnant particles in the present study, we cannot

address this question. It should furthermore be considered that gender-specific

studies on the relation between ppTG and cardiovascular risk are scarce. The above

mentioned Atherosclerosis Risk in Communities Study reported an association

between ppTG and cIMT in both men and women, but an association with exercise-

induced myocardial ischemia was only present in men and not in women [8,31].

Because of the increased risk for CVD converted by DM2 in postmenopausal

women, studies in older women are urgently needed.

Our fat-rich meal should be considered as a mixed meal because it contained

carbohydrates, through which beta-cell stimulation will be induced. Our data show

CHAPTER 3

58

that despite differences in test-meal composition and carbohydrate content (56 g

or 162 g) an association between average ppGL levels and cIMT can be found. We

also found that measuring glucose after one meal was enough to assess the

association between ppGL and cIMT. However, glucose measured two hours after

an OGTT, did not give a similar strong association with cIMT in women with NGM.

This could not be due to limited precision of the single glucose measurement after

the OGTT compared to the multiple postprandial glucose measurements, because

we found a single 2-hour glucose measurement after a meal, although borderline

significant, associated with a relevant increase in cIMT. This leads us to the

conclusion that ppGL following a meal gives additional clinical relevant information

compared to 2-hour post-load glucose following a glucose solution. First of all, we

should consider that the OGTT contained 75 gram glucose whereas the

carbohydrate-rich meal contained 162 gram carbohydrates which takes longer to

digest due to higher carbohydrate load and the more complex molecular structure

of carbohydrates in the meals. Secondly, it has been suggested that the relation

between hyperglycemia and cardiovascular disease is mediated by oxidative stress.

Ceriello et al. showed a cumulative effect of postprandial hyperglycemia and

postprandial hypertriglyceridemia on oxidative stress generation [32]. Although a

cumulative effect was possible after the mixed meal, we cannot confirm this

because we did not find an independent association between ppTG and cIMT.

Carotid IMT is a well-accepted marker for atherosclerosis and cardiovascular risk

[33]. Age and systolic blood pressure were associated with cIMT in our study. The

estimated age-related increase of 50 µm per ten years is similar to that observed in

a European study population [34] and is comparable to the age-effect in a study

among women of the same age [35]. Probably due to limited power, other CVD risk

factors, systolic blood pressure excepted, were not statistically significantly

associated with cIMT. Although the direction of the associations with cIMT were as

expected, it should be noticed that not only atherosclerosis but also other

mechanisms (thrombosis) play a role in development of CVD.

The cross-sectional study design is a potential limitation to our study. We assumed

that cIMT reflects the cumulative effect of previous atherogenic processes.

However, cIMT is a result of dynamic processes over time: progression as well as

regression due to treatment or lifestyle interventions occurs [24]. Especially in the

present population of women with DM2, who were treated intensively, the

treatment-related modification of the cIMT might have led to an underestimation of

the true association between metabolic responses and atherosclerosis.


59

The relative paucity of data in women and the disproportionately high risk of CVD

in the present study population prompted us to perform the study. However, we

should be careful in extrapolating the present results to other populations. In

addition, the present study population cannot be viewed as a population-based

cohort of post-menopausal women because from the 24% who agreed to

participate, 40% did not meet the inclusion-criteria or dropped out of the study.

The association of ppGL with cIMT in women with NGM was stronger than that of

ppTG. The association between ppGL and cIMT in women with NGM suggests that

ppGL, in the normal range, is a marker or a risk factor of CVD. Postprandial glucose

measured after a fat-rich or a carbohydrate-rich meal, might give more information

about development of atherosclerosis than 2-hour glucose, measured after an

OGTT.

ACKNOWLEDGEMENTS

This research was financially supported by a grant from the Dutch Diabetes

Research Foundation (grant no.2001.00.052) and by funding from Novartis

International AG, Switserland. The authors thank Jannet Entius, Tanja Nansink, Lida

Ooteman, Marianne Veeken and Jolanda Bosman for excellent assistance in data

collection and organization of the study, and all participants for their contributions.

CHAPTER 3

60

REFERENCES

1. Huxley R, Barzi F, Woodward M. Excess risk of fatal coronary heart disease associated

with diabetes in men and women: meta-analysis of 37 prospective cohort studies. BMJ

2006;332:73-8.

2. Yusuf S, Hawken S, Ounpuu S et al. Effect of potentially modifiable risk factors

associated with myocardial infarction in 52 countries (the INTERHEART study): case-

control study. Lancet 2004;364:937-52.

3. Turner RC, Millns H, Neil HA et al. Risk factors for coronary artery disease in non-


(UKPDS: 23). BMJ 1998;316:823-8.


85.



1999;42:926-31.

6. Hanefeld M, Koehler C, Schaper F, et al. Postprandial plasma glucose is an

independent risk factor for increased carotid intima-media thickness in non-diabetic

individuals. Atherosclerosis 1999;144:229-35.

7. DECODE Study Group. Glucose tolerance and cardiovascular mortality: comparison of

fasting and 2-hour diagnostic criteria. Arch Intern Med 2001;161:397-405.

8. Sharrett AR, Chambless LE, Heiss G, Paton CC, Patsch W. Association of postprandial

triglyceride and retinyl palmitate responses with asymptomatic carotid artery

atherosclerosis in middle-aged men and women. The Atherosclerosis Risk in

Communities (ARIC) Study. Arterioscler Thromb Vasc Biol 1995;15:2122-9.

9. Karpe F, de Faire U, Mercuri M, et al. Magnitude of alimentary lipemia is related to

intima-media thickness of the common carotid artery in middle-aged men.

Atherosclerosis 1998;141:307-14.

10. Boquist S, Ruotolo G, Tang R, et al. Alimentary lipemia, postprandial triglyceride-rich

lipoproteins, and common carotid intima-media thickness in healthy, middle-aged

men. Circulation 1999;100:723-8.



2000;23:1401-6.

12. Ahmad J, Hameed B, Das G, Siddiqui MA, Ahmad I. Postprandial hypertriglyceridemia

and carotid intima-media thickness in north Indian type 2 diabetic subjects. Diabetes

Res Clin Pract 2005;69:142-50.

13. Mari A, Tura A, Gastaldelli A, Ferrannini E. Assessing insulin secretion by modeling in

multiple-meal tests: role of potentiation. Diabetes 2002;51 Suppl 1:S221-S6.


61



15. 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;15:539-53.

16. Mooy JM, Grootenhuis PA, de Vries H et al. Prevalence and determinants of glucose


1995;18:1270-3.



Epidemiol 2003;56:1163-9.

18. Henry RM, Kostense PJ, Spijkerman AM, et al. Arterial stiffness increases with

deteriorating glucose tolerance status: the Hoorn Study. Circulation 2003;107:2089-

95.

19. Mohan V, Ravikumar R, Shanthi RS, Deepa R. Intimal medial thickness of the carotid

artery in South Indian diabetic and non-diabetic subjects: the Chennai Urban

Population Study (CUPS). Diabetologia 2000;43:494-9.

20. Voorlichtingsbureau voor de Voeding. NEVO Table Netherlands (Dutch Nutrient

Database), 2001.







23. Temelkova-Kurktschiev T, Henkel E, Schaper F, et al. Prevalence and atherosclerosis

risk in different types of non-diabetic hyperglycemia. Is mild hyperglycemia an

underestimated evil? Exp Clin Endocrinol Diabetes 2000;108:93-9.

24. Esposito K, Giugliano D, Nappo F, Marfella R. Regression of carotid atherosclerosis by

control of postprandial hyperglycemia in type 2 diabetes mellitus. Circulation 2004;

110:214-9.



1999;19:2737-41.

26. Goldschmid MG, Barrett-Connor E, Edelstein SL, et al. Dyslipidemia and ischemic heart

disease mortality among men and women with diabetes. Circulation 1994;89:991-7.

27. Kannel WB, Wilson PW. Risk factors that attenuate the female coronary disease

advantage. Arch Intern Med 1995;155:57-61.

28. Hughes TA, Elam MB, Applegate WB, et al. Postprandial lipoprotein responses in

hypertriglyceridemic subjects with and without cardiovascular disease. Metabolism

1995;44:1082-98.

CHAPTER 3

62

29. Schaefer EJ, Audelin MC, McNamara JR, et al. Comparison of fasting and postprandial

plasma lipoproteins in subjects with and without coronary heart disease. Am J Cardiol

2001;88:1129-33.

30. Karpe F, Boquist S, Tang R, et al. Remnant lipoproteins are related to intima-media

thickness of the carotid artery independently of LDL cholesterol and plasma

triglycerides. J Lipid Res 2001;42:17-21.

31. Ginsberg HN, Jones J, Blaner WS, et al. Association of postprandial triglyceride and

retinyl palmitate responses with newly diagnosed exercise-induced myocardial

ischemia in middle-aged men and women. Arterioscler Thromb Vasc Biol

1995;15:1829-1838.





33. Cheng KS, Mikhailidis DP, Hamilton G, Seifalian AM. A review of the carotid and

femoral intima-media thickness as an indicator of the presence of peripheral vascular

disease and cardiovascular risk factors. Cardiovasc Res 2002;54:528-38.

34. Kozakova M, Ferrannini E, and the RISC Investigators. Relationship between insulin

sensitivity and carotid artery intima-media thickness in healthy subjects: The RISC

Study. Diabetologia 2005;48:A-222 (Abstract).

35. Ando F, Takekuma K, Niino N, Shimokata H. Ultrasonic evaluation of common carotid

intima-media thickness (IMT)--influence of local plaque on the relationship between

IMT and age. J Epidemiol 2000; 10(1 Suppl):S10-S17.

63

Chapter 4

FASTING AND POSTPRANDIAL GLYCOXIDATIVE AND LIPOXIDATIVE STRESS ARE INCREASED IN WOMEN WITH TYPE 2 DIABETES MELLITUS


Marjan Alssema2

Peter G. Scheffer3

Michaela Diamant1


Rob Barto3

Giel Nijpels2,4

Piet J. Kostens2,5

Robert. J. Heine1,2

Casper G. Schalkwijk6

Tom Teerlink3

From the Departments of 1Endocrinology/Diabetes Center, 3Clinical Chemistry,

4General Practice and 5Clinical Epidemiology and Biostatistics, and the 2EMGO

Institute, VU University Medical Center, Amsterdam, The Netherlands, and from the

6Department of Internal Medicine, Academic Hospital Maastricht, Maastricht, The

Netherlands.

Submitted

CHAPTER 4

64

ABSTRACT

Objective: We studied acute changes in markers of glycoxidative and lipoxidative

stress, including oxidized low-density-lipoprotein, Nε-(carboxyethyl)-lysine (CEL),

Nε-(carboxymethyl)-lysine (CML) and 3-deoxyglucosone (3DG), following two

consecutive meals.

Research Design and Methods: Post-menopausal women (27 with normal glucose

metabolism [NGM]; 26 with type 2 diabetes [DM2]), received two consecutive fat-

rich meals and two consecutive carbohydrate-rich meals on two occasions. Glucose

and triglyceride concentrations were measured at baseline and 1,2,4,6 and 8h

following breakfast; lunch was given at 4h. Oxidized-LDL-to-LDL-cholesterol ratio

(oxLDL/LDL-C), CEL, CML and 3DG were measured at baseline and at 8h.

Results: Fasting oxLDL/LDL-C, 3DG and CML were higher in women with DM2

compared to women with NGM and were comparable to the postprandial values at

8 hours in NGM. Postprandial rises in oxLDL/LDL-C and 3DG were similar in both

groups. However, oxLDL/LDL-C increased more after the fat-rich meals, whereas

CML and 3DG increased more after the carbohydrate-rich meals. After the fat-rich

meals, the increase in oxLDL/LDL-C was correlated with postprandial triglycerides,

whereas the increase in 3DG was correlated with postprandial glucose.

Conclusions: The acute changes in markers of glycoxidative and lipoxidative

stress in both DM2 and NGM suggest that post-absorptive oxidative stress may

partly underlie the association of postprandial derangements and cardiovascular

risk.

POSTPRANDIAL OXIDATIVE STRESS

65

INTRODUCTION

Patients with type 2 diabetes mellitus (DM2) have an increased risk of

cardiovascular disease (CVD)[1], which can only partly be explained by the classical

CVD risk factors such as hypertension, high LDL-cholesterol, low HDL-cholesterol

and smoking [2]. In post-menopausal women, as compared to men, the relative risk

of CVD conferred by DM2 is even higher [3]. Zilversmit postulated in 1979 that

disturbances in postprandial metabolism may contribute to the excess risk of CVD

[4], due to postprandial elevations of glucose and triglyceride-enriched lipoproteins

[5,6].

Oxidative stress is regarded as a common pathway by which many of the classical

CVD risk factors and postprandial dysmetabolism may initiate and promote

atherosclerosis [7]. Indeed, elevated levels of oxidized LDL (oxLDL) are associated

with an increased risk for CVD [8]. Prolonged exposure to a high-fat diet has been

shown to result in an increase in plasma levels of oxLDL [9]. Another mechanism

that might link postprandial dysmetabolism and the risk of CVD in patients with

DM2 includes the formation of advanced glycation endproducts (AGEs), which are

related to micro- and macrovascular complications [10]. Two of the most studied

AGEs, Nε-(carboxyethyl)lysine (CEL) and Nε-(carboxymethyl)lysine (CML), can be

formed on proteins by both glycoxidation and lipid-peroxidation pathways. α-

Dicarbonyl compounds such as 3-deoxyglucosone (3DG), glyoxal and

methylglyoxal, are reactive intermediates in the formation of AGEs [11]. 3DG is a

relatively stable intermediate, whereas glyoxal and methylglyoxal convert readily

into CML and CEL, respectively. Because the latter are stable compounds, their

quantification may yield more relevant information than quantification of their

reactive dicarbonyl precursors.

Data on acute meal-induced changes in oxLDL, 3DG and CML and CEL are limited.

To date, one study in healthy young men demonstrated an increase in the oxLDL-

to-LDL-cholesterol-ratio after two consecutive fat-rich mixed meals [12], and one

study in patients with coronary artery disease (CAD) found that postprandial oxLDL

was associated with the severity and extent of CAD [13].

In the present study, 26 women with DM2 and 27 women with normal glucose

metabolism (NGM) received two consecutive (breakfast and lunch) fat-rich or

carbohydrate-rich meals, in order to study meal-related changes in oxidative stress

(oxLDL), AGE-compounds (CML and CEL), and a reactive AGE-precursor (3DG). In

CHAPTER 4

66

addition to the effects of meal composition and the presence of diabetes, we also

studied associations between postprandial changes of these markers and changes

in glucose and triglycerides.

METHODS

Subjects

The present study was a sub-study of a cross-sectional study to assess the effect

and relative contributions of two consecutive fat-rich and carbohydrate-rich meals

on markers of CVD risk in women with NGM (fasting glucose<6.1 mmol/l and 2-h

post-load glucose<7.8 mmol/l) (n=76) and DM2 (n=79). Women with DM2 (n=522),

recruited from the registry of the Diabetes Care System in the city of Hoorn, the

Netherlands, and women who were randomly selected from the municipal registry

of Hoorn (n=541), aged 50-65 years at the beginning of the study, were invited to

participate in the study. Of these 1,063 women, 431 women were complete non-

responders, 258 women were not willing to participate and 220 women did not

meet the inclusion criteria. Inclusion criteria were: post-menopausal status, non-

smoking, no untreated endocrine disorder other than DM2, no use of short acting

insulin analogues, no use of PPAR-α and -γ agonists, no use of oral corticosteroids,

no use of hormone replacement therapy, no statin use (for NGM-subjects only),

HbA1c<9.0%, fasting cholesterol<8.0 mmol/l, fasting triglycerides<4.0 mmol/l,

systolic blood pressure<190 mmHg, no liver impairment or renal impairment. For

the present study, a sub-sample of 26 women with DM2 (no use of statins or

insulin analogues) and 27 women with NGM matched for age, were studied. All

participants gave written informed consent, and the study protocol was approved

by the ethics committee of the VU University Medical Center in Amsterdam.

Study design

The study consisted of a screening visit and two visits for the test-meals. On the

screening visit, fasting blood samples were drawn after a 12h overnight fast. Blood

pressure was measured thrice with an oscillometric blood pressure measuring

device (Collin Press-mate BP-8800, Colin, Komaki-City, Japan) after a 15-minute

supine rest. Weight and height were measured in participants wearing light clothes

only. Smoking and alcohol consumption were assessed by a questionnaire [14].


67

Test meals

Postprandial meal responses were examined following two standardized

consecutive test meals (breakfast and lunch). On one day, both test meals were

high in fat content and on the other occasion, both meals were high in

carbohydrate content. Both occasions were performed in random order. The fat-rich

meals (both breakfast and lunch) consisted of 2 croissants, 10 g of butter, 40 g of

high-fat cheese and 300 ml of high-fat milk (3349kJ; 50 g fat; 56 g carbohydrates).

The carbohydrate-rich meals (both breakfast and lunch) consisted of 2 slices of

bread, 25 g of marmalade, 30 g of cooked chicken breast, 50g of ginger bread and

300ml of drinking yogurt fortified with 45 g of soluble carbohydrates (3261kJ; 4 g

fat, 162 g carbohydrates). Both meals were eaten within 10 minutes. Apart from the

test meals and water (ad libitum), participants were not allowed to eat.

Laboratory analysis

Serum total cholesterol, HDL-cholesterol and triglycerides were measured by

enzymatic colorimetric assays (Roche, Mannheim, Germany). Fasting and

postprandial LDL-cholesterol was calculated according to the Friedewald-formula if

triglycerides <4.5 mmol/l [15]. Plasma glucose levels were determined with a

glucose oxidase method (Granutest, Merck, Darmstadt, Germany) and HbA1c was

measured with cation-exchange chromatography (Menarini Diagnostics, Florence,

Italy).

OxLDL was measured in EDTA-plasma in duplicate with a competitive ELISA

(Mercodia, Uppsala, Sweden), with inter-assay and intra-assay coefficients of

variation of 7.8% and 4.8%, respectively. We expressed oxLDL as a ratio per LDL-

cholesterol (oxLDL/LDL-C) as proposed by Holvoet et al [16].

3DG was measured with a newly developed LC-MS/MS method after pre-column

derivatization according to a published procedure [17]. In brief, 0.5 ml of whole

blood was mixed with 1.0 ml of 1.2 mol/l perchloric acid and immediately stored at

-80 ˚C until assayed. After thawing, the samples were centrifuged (5 min at 20,000

g) and 80 µl of the supernatant was mixed with 100 µl of the internal standard

solution (1 µmol/l 2,3-pentanedione dissolved in ethanol). After addition of 20 µl of

the derivatization reagent (10 mmol/l 2,4-dinitrophenylhydrazine dissolved in 1.2

mol/l perchloric acid), the samples were incubated for 16 h at room temperature.

Chromatographic separation was performed on a Xterra MS C18-column (4.6x50

CHAPTER 4

68

mm; 3.5 µm particle size) from Waters (Milford, MA, USA) using a linear gradient

from 40% to 95% acetonitrile in water over 9 min at a flow rate of 1 ml/min. Mass

transitions of 521.1→431.0 and 521.1→182.1 for 3DG and 459.1→182.1 for the

internal standard were monitored in negative-ion mode. Intra-assay and inter-assay

coefficients of variation for 3DG were 8% and 12%, respectively. Two subjects with

NGM had missing 3DG because of missing blood samples. In a pilot study applying

the same study-protocol, we assessed the time-course of 3DG in 9 DM2 and 8 NGM

subjects and measured 3DG with an HPLC-method with fluorescence detection as

described previously [18]. The highest 3DG concentration was observed at 8 hours

after the first meal [19].

Unbound CML and CEL were measured in EDTA-plasma by HPLC with stable-

isotope-dilution tandem mass spectrometric detection (LC-MS/MS) using a

procedure developed for the analysis of protein-bound CML and CEL [20], with a

modified sample preparation procedure. Briefly, after centrifugation of the plasma

samples (5 min at 20,000 g), 50 µl of the supernatant was mixed with 150 µl of a

combined internal standard solution (containing deuterated CML and CEL) and

proteins were removed by centrifugation (20 min at 15,000 g) in a 10 kDa cut-off

ultrafiltration device (Vivaspin 500 PES from Vivascience AG, Hannover, Germany).

A 90 µl aliquot of the ultrafiltrate was mixed with 10 µl of a 50 mmol/l

nonafluoropentanoic acid solution and 25 µl of this mixture was subjected to LC-

MS/MS analysis as described before [20]. Intra-assay and inter-assay coefficients of

variation for CML were 3% and 4%, respectively and for CEL 5% and 9%, respectively.

To assess the cumulative effect of both consecutive meals we measured oxLDL,

3DG, CML, and CEL at baseline and at 8h.


Analyses were performed with SPSS for Windows 11.01 (SPSS Inc., Chicago, IL, USA).

Data are presented as mean values (standard deviation), and in case of a skewed

distribution, as median values (interquartile range). Data were ln-transformed

before testing in case of a skewed distribution. Differences between women with

NGM and DM2 were tested with Student’s t-test for continuous variables and c2-test

for dichotomous variables. Differences between meals were tested with paired

samples t-tests. Postprandial changes in glucose and triglycerides were calculated

as incremental area under the curve (iAUC) by the trapezoid method. Correlations

between postprandial responses were calculated by Spearman’s correlation-

coefficients. We considered a two-sided P<0.05 to indicate statistical significance.


69

RESULTS

Characteristics of the study population

The clinical and biochemical characteristics of the participants at the screening visit

are presented in Table 4.1.

Table 4.1. Clinical and biochemical characteristics of the study-population at the screening visit Variables NGM DM2 P-value

N 27 26

Age (years) 60.3 (4.0) 60.4 (3.2) 0.93

BMI (kg/m2) 26.0 (3.1) 31.8 (5.8) <0.001

Fasting glucose (mmol/l) 5.5 (0.3) 7.6 (1.3) <0.001

HbA1c (%) 5.6 (0.3) 6.5 (0.6) <0.001

Total cholesterol (mmol/l) 6.0 (1.0) 5.6 (0.9) 0.69

HDL-cholesterol (mmol/l) 1.93 (0.57) 1.47 (0.30) 0.01

LDL-cholesterol (mmol/l) 3.6 (0.9) 3.3 (0.9) 0.22

Triglycerides (mmol/l)* 0.9 (0.8-1.6) 2.0 (1.4-2.4) <0.001

Systolic blood pressure (mmHg) 125 (12) 143 (16) <0.001

Diastolic blood pressure (mmHg) 69 (8) 79 (7) <0.001

Smoking (former) 56% 39% 0.16

Alcohol-intake (g/week) 78 ± 12 24 ± 7 <0.01

Antihypertensive medication (%) 0 58 <0.001

Diabetes treatment (%)

- diet alone - 35 -

- metformin - 54 -

- sulfonylureas - 19 -

- acarbose - 4 -

Data are mean (standard deviation) or median (interquartile range) in case of skewed data NGM= normal glucose metabolism, DM2= type 2 diabetes mellitus *ln-transformed before analysis

CHAPTER 4

70

Figure 4.1. Time course of triglycerides and glucose concentrations (mean±SD) following

two consecutive fat-rich or carbohydrate-rich meals in women with normal glucose

metabolism (open circles) and type 2 diabetes mellitus (black circles).

Figure 4.1 shows the 8 h-time courses of triglycerides and glucose after two

consecutive fat-rich and after two consecutive carbohydrate-rich meals. At both

meal visits, fasting triglyceride and glucose levels were higher in women with DM2

compared to the women with NGM (both P<0.001). Triglycerides-iAUC was higher

following the fat-rich meals than after the carbohydrate-rich meals in both women

with DM2 and NGM (both P<0.001), but was similar in both groups. Following the

carbohydrate-rich meals, glucose-iAUC was higher in women with DM2 compared to

NGM (P<0.001), whereas this difference was borderline significant after the fat-rich

meals (P=0.06).

LDL-C significantly decreased from baseline to 8h in both groups and after both

meal types (from 3.6±0.8 mmol/l to 3.2±0.7 mmol/l and 3.7±0.8 mmol/l to

3.3±0.7 mmol/l, fat-rich meals and carbohydrate-rich meals, respectively in NGM

and from 3.3±0.7 mmol/l to 2.9±0.7 mmol/l and 3.3±0.7 mmol/l to 3.0±0.7


71

mmol/l, fat-rich meals and carbohydrate-rich meals, respectively in DM2 (all

P<0.001).

Glycoxidative and lipoxidative stress markers

The oxLDL/LDL-C, 3DG, CML and CEL values in women with NGM and DM2 at

baseline and at 8h are listed in Table 2. Baseline oxLDL/LDL-C, 3DG and CML were

significantly higher in DM2 compared to NGM. The increase of oxLDL/LDL-C during

the fat-rich meals was higher than the increase during the carbohydrate-rich meals

(17±10% versus 10±7% in NGM and 15±13% versus 7±9% in DM2, both P<0.05). Of

interest, the fasting values of ox-LDL/LDL-C, 3DG and CML in women with DM2

were similar to the postprandial values at 8h in women with NGM (all P>0.1).

Table 4.2. Fasting and postprandial (8 hours) levels of oxLDL/LDL-C, 3DG, CML and CEL in

women with NGM compared to women with DM2*

NGM DM2

Fasting Postprandial Fasting Postprandial

OxLDL/LDL-C (U/mmol) Fat 21.6 (3.8) 25.1 (4.8)‡ 25.4 (4.6)‡ 29.6 (8.7)‡

Carb 21.6 (2.8) 23.7 (3.8)‡ 25.7 (4.3)‡ 27.3 (4.9)‡

3DG (nmol/l) Fat 164 (41) 183 (45)‡ 208 (60)§ 236 (58)‡

Carb 164 (37) 208 (40)§ 210 (49)§ 281 (50)§

CML (nmol/l) Fat 39.6 (10.1) 45.1 (10.6)‡ 52.2 (18.1)‡ 47.9 (14.1)†

Carb 43.3 (10.7) 54.7 (9.4)§ 48.2 (16.2)‡ 55.9 (16.0)§

CEL (nmol/l) Fat 62.7 (31.2) 67.4 (15.1) 67.0 (27.4) 66.0 (16.3)

Carb 56.4 (13.5) 72.1 (12.8) § 65.8 (23.3) 72.9 (17.9)

*Abbrevations: NGM:normal glucose metabolism, DM2:type 2 diabetes mellitus,

Fat:fat-rich meal, Carb:carbohydrate-rich meal, OxLDL/LDL-C:ratio of oxidized LDL to total LDL-cholesterol,

3DG:3-dexoxyglucosone, CML:Nepsilon-(carboxymethyl)-lysine, CEL:Nepsilon-(carboxyethyl)-lysine

†p<0.05 DM2 baseline versus NGM baseline or 8h versus baseline

‡p<0.01 DM2 baseline versus NGM baseline or 8h versus baseline

§p<0.001 DM2 baseline versus NGM baseline or 8h versus baseline

Baseline oxLDL/LDL-C was correlated with fasting triglycerides in women with DM2

(r=0.39, P<0.05), but not in women with NGM. After the fat-rich meals, changes in

oxLDL/LDL-C correlated with changes in triglycerides, either expressed as

triglycerides-iAUC (r=0.32, P<0.05; Figure 4.2a) or as increase over baseline at 8h

(r=0.50, P<0.001).

CHAPTER 4

72

The mean increase of 3DG after the carbohydrate-rich meals was higher than after

the fat-rich meals (29±21% versus 13±24%, P<0.05 in NGM, and 36±17% versus

15±20%, P<0.001 in DM2). Fasting plasma glucose concentration was correlated

with baseline 3DG in women with DM2 (r=0.50, P<0.01), but not in women with

NGM. Postprandial changes in 3DG correlated to glucose-iAUC after the fat-rich

meals (r=0.37, P<0.01; Figure 4.2b), but not with glucose-iAUC after the

carbohydrate-rich meals. The changes of CML and CEL were neither correlated to

triglycerides-iAUC nor to glucose-iAUC, irrespective of meal composition.

Figure 4.2. Relations between postprandial changes of the oxLDL-to-LDL-cholesterol ratio

and 8h incremental area under the curve for plasma triglycerides (A) and between changes

of 3-deoxyglucosone and 8h incremental area under the curve for glucose (B) after two

consecutive fat-rich meals in women with normal glucose metabolism (open circles) and

type 2 diabetes mellitus (black circles). Strengths of the associations are indicated by

Spearman’s correlation coefficients.


73

DISCUSSION

In the present cross-sectional study in postmenopausal women with NGM and DM2,

we studied the postprandial changes of oxLDL/LCL-C, 3DG, CML and CEL, following

two consecutive fat-rich or carbohydrate-rich meals. The main findings are that the

fasting values of oxLDL/LDL-C, 3DG and CML are higher in women with DM2

compared to women with NGM. In women with DM2, the fasting values of

oxLDL/LDL-C, 3DG and CML were similar to the postprandial values at 8h in women

with NGM. Overall, the postprandial rises of oxLDL/LDL-C and 3DG were of similar

magnitude in DM2 and NGM. OxLDL/LDL-C showed a greater increase after the fat-

rich meals, whereas CML and 3DG increased more after the carbohydrate-rich

meals. The increase in oxLDL/LDL-C was correlated with the rise in postprandial

triglycerides, whereas the increase in 3DG was correlated with the postprandial rise

of glucose.

There is increasing evidence that postprandial glucose and triglycerides may play

an important role in macro- and microvascular complications [21]. However, their

relative contributions and the mechanisms leading to the increased CVD risk are

not yet fully understood [5]. Prospective studies have shown that postprandial or

post-load glucose levels are more strongly associated with increased CVD than

fasting glucose [22,23]. Oxidative stress is hypothesized to be an important

pathway linking postprandial glucose and triglyceride responses to CVD [24,25].

Indeed, in a recent study it was found that postprandial oxLDL was a determinant

of the extent of coronary atherosclerosis in patients with CAD [13]. A limitation of

most previous studies is that responses after a single meal were studied, which

might not reflect the “real life” burden of the postprandial state, possibly leading to

an underestimation of the associations between postprandial responses and CVD

risk. A recent study from our group that applied two consecutive meals (breakfast

and lunch) in healthy young men, demonstrated that even slightly increased levels

of triglycerides and glucose were associated with endothelial dysfunction and

increased markers of oxidative stress, including oxLDL [12]. We concluded that

these findings might be relevant for insulin resistant subjects and DM2 individuals

who have a more pronounced (exaggerated and prolonged) postprandial elevation

of glucose and triglycerides, possibly leading to more oxidative stress. Our present

findings support this conjecture, as we found that baseline oxLDL was higher in

DM2 than in NGM, with a greater increase of oxLDL after the fat-rich meals

compared to the carbohydrate-rich meals. Interestingly, although fasting

CHAPTER 4

74

oxLDL/LDL-C was higher in the DM2 group, we found no difference in the

postprandial increase in oxLDL/LDL-C between NGM and DM2.

There are multiple sources and mechanisms in the formation of AGEs in vivo,

involving both oxidative and non-oxidative reactions of reducing carbohydrates and

other metabolic intermediates with amino acid residues [26]. The AGE precursor

3DG, which is increased in patients with diabetes [27], is formed by non-oxidative

modifications of Amadori adducts and from fructose-3-phosphate [28]. In line with

these observations, we found that 3DG increased more after the carbohydrate-rich

meals than after the fat-rich meals. In addition, postprandial changes in 3DG were

strongly correlated to changes in glucose, but surprisingly, only after the fat-rich

meals, suggesting additional involvement of oxidative mechanisms in its formation.

In patients with type 1 diabetes Beisswenger et al observed an increase in 3DG

within two hours [29], whereas in our pilot study in patients with DM2 we did not

observe a rise in 3DG after the first meal [19].

Some studies have shown that total serum CML is higher in DM2 than in healthy

individuals [30,31], although in other studies no difference was observed [20]. We

chose not to measure total CML and CEL, but rather the unbound forms of these

compounds, that are continuously released during degradation of intracellular

proteins with a rapid turn-over. Because oxidative stress not only enhances

formation of protein-bound CML and CEL, but also triggers proteasomal

degradation of damaged proteins [32,33], the unbound fraction of CML and CEL

may better reflect formation of AGEs by oxidative stress. We observed that CML was

affected by both meal types. However, possibly due to a relatively large biological

variation, as reflected by differences in the baseline CML and CEL values at the two

meal visits, these changes were less consistent than for oxLDL and 3DG. Future

studies should assess whether changes in the protein-bound forms of CML and CEL

are more informative.

The major precursor of CML is glyoxal and oxidative stress seems to enhance the

formation of glyoxal from glucose [34], which is consistent with our finding of a

greater increase in CML after the carbohydrate-rich meals compared to the fat-rich

meals.


75

Potential limitations

First, both groups were not matched for BMI, lipids, blood pressure and alcohol

intake, all of which may have independent effects on the parameters measured.

However, we chose not to match for these variables, to avoid selection of relatively

healthy DM2 patients. Second, our study was performed in elderly postmenopausal

women and the changes observed may be different in younger women as well as in

men.

Conclusion

In summary, we found significant elevations of postprandial oxLDL/LDL-C, 3DG and

CML in postmenopausal women with DM2 and NGM. Since postprandial increases in

α-dicarbonyls and associated AGEs are related to cellular damage, therapies aimed

at reducing postprandial glycemia and triglyceridemia, may contribute to the

reduction in micro- and macrovascular complications.

ACKNOWLEDGEMENTS

The Hoorn prandial study was financially supported by funding from Novartis AG

Switzerland and by a grant from the Dutch Diabetes Research Foundation (grant

no.2001.00.052). We thank Danielle van Assema for her assistance in organizing

the pilot-study and Rick Vermue for his excellent technical assistance.

CHAPTER 4

76

REFERENCES

1. Haffner SM, Lehto S, Ronnemaa T, et al. Mortality from coronary heart disease in

subjects with type 2 diabetes and in nondiabetic subjects with and without prior


2. Turner RC, Millns H, Neil HA, et al. Risk factors for coronary artery disease in non-


(UKPDS: 23). BMJ 1998;316:823-8.





85.



6. Tushuizen ME, Diamant M, Heine RJ. Postprandial dysmetabolism and cardiovascular

disease in type 2 diabetes. Postgrad Med J 2005;81:1-6.

7. Haidara MA, Yassin HZ, Rateb M, et al. Role of oxidative stress in development of

cardiovascular complications in diabetes mellitus. Curr Vasc Pharmacol 2006;4:215-

27.

8. Meisinger C, Baumert J, Khuseyinova N, et al. Plasma oxidized low-density

lipoprotein, a strong predictor for acute coronary heart disease events in apparently

healthy, middle-aged men from the general population. Circulation 2005;112:651-7.

9. Silaste ML, Rantala M, Alfthan G, et al. Changes in dietary fat intake alter plasma

levels of oxidized low-density lipoprotein and lipoprotein(a). Arterioscler Thromb

Vasc Biol 2004;24:498-503.

10. Kilhovd BK, Berg TJ, Birkeland KI, et al. Serum levels of advanced glycation end

products are increased in patients with type 2 diabetes and coronary heart disease.

Diabetes Care 1999;22:1543-8.

11. Kato H, Shin DB, Hayase F. 3-Deoxyglucosone Cross-Links Proteins Under

Physiological Conditions. Agricultural and Biological Chemistry 1987;51:2009-11.




13. Graner M, Kahri J, Nakano T, et al. Impact of postprandial lipaemia on low-density

lipoprotein (LDL) size and oxidized LDL in patients with coronary artery disease. Eur J

Clin Invest 2006;36:764-70.



1995;18:1270-3.


77




16. Holvoet P, Kritchevsky SB, Tracy RP, et al. The metabolic syndrome, circulating

oxidized LDL, and risk of myocardial infarction in well-functioning elderly people in

the health, aging, and body composition cohort. Diabetes 2004;53:1068-73.

17. Zwiener C, Glauner T, Frimmel FH. Method optimization for the determination of

carbonyl compounds in disinfected water by DNPH derivatization and LC-ESI-MS-MS.

Anal Bioanal Chem 2002;372:615-21.

18. Schalkwijk CG, Posthuma N, Ten Brink HJ, et al. Induction of 1,2-dicarbonyl

compounds, intermediates in the formation of advanced glycation end-products,

during heat-sterilization of glucose-based peritoneal dialysis fluids. Perit Dial Int

1999;19:325-33.

19. Schindhelm RK, Alssema M, Heine RJ, et al. Carbohydrate, but not fat-rich meals

cause a rise in plasma 3-deoxyglucosone concentration in females with type 2

diabetes mellitus [abstract]. Diabetes 2005;54:A163.

20. Teerlink T, Barto R, Ten Brink HJ, Schalkwijk CG. Measurement of Nepsilon-

(carboxymethyl)lysine and Nepsilon-(carboxyethyl)lysine in human plasma protein by

stable-isotope-dilution tandem mass spectrometry. Clin Chem 2004;50:1222-8.





1999;42:926-31.

23. 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 Diagnostic

criteria in Europe. Lancet 1999;354:617-21.





25. Anderson RA, Evans ML, Ellis GR, et al. The relationships between post-prandial

lipaemia, endothelial function and oxidative stress in healthy individuals and patients

with type 2 diabetes. Atherosclerosis 2001;154:475-83.

26. Baynes JW, Thorpe SR. Role of oxidative stress in diabetic complications: a new

perspective on an old paradigm. Diabetes 1999;48:1-9.

27. Hamada Y, Nakamura J, Fujisawa H, et al. Effects of glycemic control on plasma 3-

deoxyglucosone levels in NIDDM patients. Diabetes Care 1997;20:1466-9.

CHAPTER 4

78

28. Thornalley PJ, Langborg A, Minhas HS. Formation of glyoxal, methylglyoxal and 3-

deoxyglucosone in the glycation of proteins by glucose. Biochem J 1999;344 Pt

1:109-16.

29. Beisswenger PJ, Howell SK, O'Dell RM, et al. alpha-Dicarbonyls increase in the

postprandial period and reflect the degree of hyperglycemia. Diabetes Care

2001;24:726-32.

30. Schleicher ED, Wagner E, Nerlich AG. Increased accumulation of the glycoxidation

product nepsilon -(carboxymethyl)lysine in human tissues in diabetes and aging. J

Clin Invest 1997;99:457-68.

31. Wautier MP, Massin P, Guillausseau PJ, et al. N(carboxymethyl)lysine as a biomarker

for microvascular complications in type 2 diabetic patients. Diabetes Metab

2003;29:44-52.

32. Shringarpure R, Grune T, Mehlhase J, Davies KJ. Ubiquitin conjugation is not required

for the degradation of oxidized proteins by proteasome. J Biol Chem 2003;278:311-

8.

33. Mehlhase J, Grune T. Proteolytic response to oxidative stress in mammalian cells. Biol

Chem 2002;383:559-67.

34. Nagai R, Unno Y, Hayashi MC, et al. Peroxynitrite induces formation of N( epsilon )-

(carboxymethyl) lysine by the cleavage of Amadori product and generation of

glucosone and glyoxal from glucose: novel pathways for protein modification by

peroxynitrite. Diabetes 2002;51:2833-9.

79

Chapter 5

COMPARISON OF THE EFFECT OF TWO CONSECUTIVE FAT-RICH AND CARBOHYDRATE-RICH MEALS ON POSTPRANDIAL PLASMA

MYELOPEROXIDASE IN WOMEN WITH AND WITHOUT TYPE 2 DIABETES MELLITUS


Marjan Alssema2

Michaela Diamant1

Tom Teerlink3


Astrid Kok3

Piet J. Kostens2,5

Giel Nijpels2,4

Robert. J. Heine1,2

Peter G. Scheffer3

From the Departments of 1Endocrinology/Diabetes Center, 3Clinical Chemistry

4General Practice and 5Clinical Epidemiology and Biostatistics, and the 2EMGO

Institute, VU University Medical Center, Amsterdam, The Netherlands.

Submitted

CHAPTER 5

80

ABSTRACT

Background: Patients with type 2 diabetes mellitus (DM2) have an increased risk of

cardiovascular disease (CVD). Myeloperoxidase (MPO), expressed in leukocytes, and

released upon activation, is associated with CVD and endothelial dysfunction.

Postprandial leukocyte recruitment and activation with subsequent MPO release

may contribute to atherosclerosis and CVD. We hypothesized that MPO may

increase in the postprandial state, due to postprandial leukocyte recruitment

and/or activation, especially in subjects with DM2.

Methods: One hundred post-menopausal women, aged 50-65 years (66 with

normal glucose metabolism [NGM], and 34 with DM2), received two consecutive fat-

rich meals and two consecutive carbohydrate-rich meals on separate occasions.

Blood samples were taken before (t=0), and at 2,4 and 8 hours following breakfast;

lunch was given at t=4. Plasma MPO-concentration was measured by sandwich

ELISA.

Results: The number of leukocytes in fasting blood samples was higher in DM2

compared with NGM (6.1±1.4 and 5.4±1.2x109/l, respectively; P<0.05). Baseline

MPO concentration did not differ between NGM and DM2 (51.4±12.9 and

54.5±18.4 µg/L, respectively; P=0.39). Baseline MPO was positively associated with

leukocytes (r=0.20; P<0.05) and inversely associated with high-density lipoprotein

cholesterol (r=-0.22; P<0.05). Leukocytes increased from 5.0±1.5 to 6.1±1.5x109/L

and from 5.8±1.4 to 6.6±1.4x109/l in NGM and DM2, respectively (both P<0.01)

following the fat-rich meals. In contrast to our hypothesized increase in MPO, we

found a significant decrease in MPO in NGM (both meal types) and DM2 (fat-rich

meals only).

Conclusions: Our findings provide no support to our initial hypothesis that meal-

induced release of MPO might be a mechanism that contributes to CVD risk.

POSTPRANDIAL MYELOPEROXIDASE

81

INTRODUCTION

Patients with type 2 diabetes mellitus (DM2) have an increased risk of

cardiovascular disease (CVD) [1-3[. Evidence has been accumulating that chronic

low-grade inflammation may be an important factor in the pathogenesis of

atherosclerosis and CVD [4,5]. The number of white blood cells, especially

polymorphonuclear neutrophils, is associated with future CVD events [6,7].

However, the role of leukocytes in the pathogenesis of CVD has not yet been fully

elucidated. It has been demonstrated that the number of leukocytes was increased

between one and six hours after a fat challenge [8]. Postprandial recruitment and

activation of these cells have been suggested to contribute to endothelial

dysfunction [9], which is an early event in the atherosclerotic process. Since

postprandial responses after a fat-rich mixed-meal may be more pronounced and

prolonged in subjects with DM2 than in healthy controls [10], this may result in a

higher postprandial leukocyte recruitment in these patients. This may be even more

the case in postmenopausal women, who have higher postprandial triglyceride

responses compared to premenopausal women [11].

Myeloperoxidase (MPO) is a heme protein which is abundantly expressed in

leukocytes, and released upon activation into phagocytic vacuoles and in the extra

cellular space [12,13]. Several in vitro and in vivo studies have shown that MPO

interacts with the vessel-wall by various mechanisms, including binding and

transcytosis through endothelial cells, and leads to hypochlorous acid generation,

nitric oxide oxidation and tyrosine nitration [14]. MPO has been associated with

coronary artery disease [15], recurrent coronary events [16], heart failure [17] and

endothelial dysfunction [18].

Postprandial leukocyte recruitment and activation with subsequent MPO activation

may thus contribute to atherosclerosis and CVD, especially in patients with DM2

because of the prolonged and exaggerated dysmetabolic state. Based on these

observations, we hypothesized that, in particular in subjects with DM2, MPO would

increase in the postprandial state due to postprandial leukocyte recruitment and/or

activation. Therefore we assessed postprandial plasma MPO after two consecutive

fat-rich or carbohydrate-rich meals in women with DM2 and in women with normal

glucose metabolism (NGM).

CHAPTER 5

82

MATERIAL AND METHODS

Subjects and study design

The women included in the present study were part of a cross-sectional study to

assess the effect and relative contributions of two consecutive fat-rich and

carbohydrate-rich meals on markers of CVD risk in postmenopausal women with

NGM (n=76) and DM2 (n=79). Women with DM2 (n=522), recruited from the registry

of the Diabetes Care System in the city of Hoorn, the Netherlands, and women who

were randomly selected from the municipal registry of Hoorn (n=541), aged 50-65

years at the beginning of the study, were invited to participate in the study. Of

these 1,063 women, 431 women were complete non-responders, 258 women were

not willing to participate and 219 women did not meet the inclusion criteria,

yielding 155 eligible participants. Inclusion criteria were: post-menopausal status

(no menses in the last 12 months), non-smoking, no untreated endocrine disorder

other than DM2, no use of short acting insulin analogues, no use of PPAR-α and -γ

agonists, no use of oral corticosteroids, no use of hormone replacement therapy,

no use of HMG-CoA reductase inhibitors (statins) (for NGM only), HbA1c<9.0%,

fasting cholesterol <8.0 mmol/l, fasting triglycerides <4.0 mmol/l, systolic blood

pressure <190 mmHg, no liver impairment or renal impairment. Women who were

selected from the municipal registry underwent a 75-g oral glucose tolerance test

to verify their glucose tolerance status (fasting glucose <6.1 mmol/l and 2-hour

post-load glucose <7.8 mmol/l). In addition to the above-mentioned inclusion

criteria of the main study protocol, for the present analyses we chose not to include

DM2 women using statins (n=38), as statins have been shown to down regulate

MPO gene expression [19,20]. Thus, for the present study 41 women with DM2,

and 76 women with NGM were included. The protocol consisted of three separate

visits, i.e. a screening visit and two visits for the test-meals. Postprandial meal

responses were examined following the consumption of two consecutive

standardized test meals (breakfast and lunch; fat-rich meals [3349 kJ; 50 g fat; 56

g carbohydrates and 28 g proteins] and carbohydrate meals [3261 kJ; 4 g fat, 162 g

carbohydrates and 22 g of proteins]) on two separate occasions and in random

order. Blood samples were taken before and at t=2, 4 and 8 hours following

breakfast for measurement of glucose, triglycerides and MPO. Lunch was given 4

hours after breakfast (t=4). Apart from the test meals and water (ad libitum),

participants were not allowed to eat. In addition, physical activity was limited

during the day. All participants gave informed consent, and the study protocol was


83

approved by the ethics committee of the VU University Medical Center in

Amsterdam.

Measurements

At the screening visit, blood pressure was measured at the left arm three times

with 5 minute intervals with an oscillometric blood pressure measuring device

(Collin Press-mate BP-8800, Colin, Komaki-City, Japan) after a 15-minute supine

rest. Waist circumference was measured twice at the level midway between the

lowest rib margin and the iliac crest. Medical history was assessed by a

questionnaire [21].

Laboratory analyses

MPO was determined in duplicate by a sandwich ELISA (Mercodia, Uppsala, Sweden)

in EDTA-plasma using a CODA® automated EIA analyzer (Biorad, San Francisco, CA,

USA). The concentration of MPO was measured in baseline samples, and in samples

that were taken at 2, 4 and 8 hours after breakfast. The intra-assay and inter-assay

coefficients of variation were 3.3% and 5.0%, respectively. Serum total cholesterol,

high-density lipoprotein (HDL) cholesterol and triglycerides were measured by

enzymatic colorimetric assays (Roche, Mannheim, Germany). Low-density

lipoprotein (LDL) cholesterol was calculated according to the Friedewald-formula

[22]. Plasma glucose levels were determined with a glucose oxidase method

(Granutest, Merck, Darmstadt, Germany) and glycated hemoglobin (HbA1c) was

measured with cation exchange chromatography (Menarini Diagnostics, Florence,

Italy). Leukocytes were measured at baseline in all subjects and at 8 hours after

breakfast in a sub-sample (n=43; 29 NGM and 14 DM2) to determine the

postprandial leukocyte response.


Analyses were performed in SPSS for Windows 11.0.1 (SPSS Inc. Chicago, IL). Data

are presented as mean values (standard deviation) or in case of skewed

distribution, as median values (interquartile range). Differences in baseline

characteristics between the NGM and DM2 groups were tested with a Students’ t-

test for continuous variables. Postprandial triglyceride, glucose and MPO responses

were calculated as incremental and total area under the curve (iAUC and tAUC,

respectively) with the trapezoid method and expressed as mean incremental or

total concentration by dividing the incremental or total response by 8 hours.

CHAPTER 5

84

Differences in postprandial MPO responses (mean total and mean incremental area

under the curve) between the fat-rich and the carbohydrate-rich meals were tested

by a paired samples t-test. A one-samples t-test with zero as test-value was used to

test whether the mean incremental area was different from zero (i.e. no change in

mean incremental MPO). Differences in postprandial time-point (2, 4 and 8h) were

tested with a paired-samples t-test compared to baseline. Correlations between

MPO and clinical and biochemical characteristics were expressed as Spearman’s

correlation coefficients. A two-sided P-value<0.05 was considered as statistically

significant.

RESULTS

Characteristics of study population

Subjects with one or more missing values were excluded (n=17) and all statistical

analyses were performed on the remaining 100 subjects (66 with NGM and 34 with

DM2). Subjects who were excluded from the statistical analyses did not significantly

differ with respect to fasting total leukocyte count, triglycerides and glucose (all

P>0.05, for NGM and DM2 separately). The characteristics of the 100 subjects are

presented in Table 5.1. On average, women with DM2 were more obese, had higher

fasting plasma glucose, HbA1c, triglycerides, and blood pressure compared to

women with NGM. No differences were observed in total and LDL-cholesterol,

whereas HDL-cholesterol was lower in DM2.

Postprandial glucose and triglycerides

Figure 1 shows the 8 hour time courses of triglycerides and glucose after two

consecutive fat-rich and after two consecutive carbohydrate-rich meals. The

increase (iAUC) in postprandial triglycerides was similar in women with DM2 and

women with NGM after the fat-rich meals (0.9±0.5 versus 0.9±0.5 mmol/l; P=0.97).

The rise of plasma glucose (iAUC) levels after the fat-rich meals was higher in

women with DM2 compared to the NGM women (0.7±1.5 versus 0.1±0.4 mmol/l,

respectively; P<0.05). Following the carbohydrate-rich meals, postprandial glucose

(iAUC) values were higher in women with DM2, compared to women with NGM

(3.1±2.7 versus 0.4±0.6 mmol/l, respectively; P<0.001).


85

Table 5.1. Clinical and biochemical characteristics of the study-population (n = 100)

Variablea NGM DM2 P-value

n 66 34

Age, years 60.5 (4.0) 58.7 (3.7) 0.03

Waist, cm 87.1 (9.8) 102.4 (14.4) <0.001

Fasting glucose, mmol/lb 5.2 (0.3) 7.0 (1.3) <0.001

HbA1c, % 5.5 (0.3) 6.5 (0.6) <0.001

Total cholesterol, mmol/lb 5.7 (0.8) 5.6 (1.0) 0.41

HDL-cholesterol, mmol/lb 1.65 (0.41) 1.40 (0.30) <0.01

LDL-cholesterol, mmol/lb 3.5 (0.8) 3.4 (0.8) 0.40

Triglycerides, mmol/lb, c 1.1 (0.9-1.4) 1.5 (1.2-2.0) <0.001

Systolic blood pressure, mmHg 129 (14) 143 (17) <0.001

Diastolic blood pressure, mmHg 70 (8) 79 (7) <0.001

NGM= normal glucose metabolism, DM2= type 2 diabetes mellitus a Data are mean (standard deviation) b mean of the two meal-visits c ln-transformed before analysis

Figure 5.1. Time course of triglycerides [left panel] and glucose [right panel] (mean ± SEM)

following two consecutive carbohydrate-rich meals (solid line) and two fat-rich meals

(dotted line), given at 0 and 4 hours in individuals with normal glucose metabolism (NGM:

open dots) and patients with type diabetes mellitus (DM2: solid dots)

0 2 4 6 8

0

5

10

15

20

time [hours]

glucose [mmol/L]

0 2 4 6 8

0

1

2

3

4

time [hours]

triglycerides[m

mol/L]

0 2 4 6 8

0

5

10

15

20

time [hours]

glucose [mmol/L]

0 2 4 6 8

0

1

2

3

4

time [hours]

triglycerides[m

mol/L]

CHAPTER 5

86

Baseline and postprandial leukocytes and MPO

Baseline number of leukocytes was higher in DM2 compared to NGM (6.1±1.4

versus 5.4±1.2 x 109 /l, respectively; P < 0.05). Changes in postprandial leukocyte

counts were measured in a sub-sample of 43 participants (29 NGM and 14 DM2). In

this sub-sample, the number of leukocytes did not significantly increase after the

carbohydrate-rich meals (from 5.0±1.2 to 5.3±1.3 x 109 /l in NGM and from

5.8±1.3 to 6.0±1.4 x 109 /l in DM2; P = 0.07 and P = 0.23, respectively). Following

the fat-rich meals, leukocytes increased from 5.0±1.5 to 6.1±1.5 x 109 /l and from

5.8±1.4 to 6.6±1.4 x 109 /l in NGM and DM2, respectively (both P < 0.01). The

postprandial changes in MPO are presented in Figure 2 and in Table 2. Baseline

MPO (mean value of the two meal visits) was not statistically significantly different

between NGM and DM2 (51.4±12.9 and 54.5±18.4 µg /l respectively; P = 0.39).

Baseline MPO was positively associated with total leukocyte counts (r=0.20; P =

0.04), whereas an inverse association of fasting MPO with fasting HDL-cholesterol

(r=-0.22; P = 0.02) was observed. No significant associations of baseline MPO were

found with age, triglycerides, fasting glucose, and total-cholesterol and systolic and

diastolic blood pressure. MPO-iAUC and MPO-tAUC after both meal types did not

significantly differ between DM2 and NGM, nor did MPO at any of the time-points

differ between DM and NGM (all P > 0.1). MPO-iAUC was significantly different from

zero in NGM (-2.5±5.5 and -2.9±5.3 µg/l, fat-rich meals and carbohydrate-rich

meals respectively; both P<0.01), but not in DM2 (-1.7±5.5 and -1.6±5.2 µg/l; fat-

rich meals and carbohydrate-rich meals respectively; P = 0.08 and P = 0.09

respectively). Furthermore, compared to baseline MPO was significantly lower at 8

hours after breakfast in NGM (both meal types) and DM2 (fat-rich meals only) (all P

< 0.05) (Table 2).

Table 5.2. Fasting and postprandial levels of MPO (mean, SD) in women with NGM and women with DM2a

Time CARB FAT

NGM DM2 NGM DM2

MPO [µg/l] baseline 51.8 (14.0) 55.6 (18.9) 51.0 (13.0) 53.8 (18.4)

2 50.6 (13.1) 52.3 (18.3) b 51.9 (13.8) 52.3 (17.4)

4 48.8 (13.4)b 54.1 (17.2) 49.5 (12.1) 53.2 (17.8)

8 48.0 (13.0)b 53.5 (18.5) 47.2 (12.6)b 50.5 (17.7)b

aAbbrevations: NGM: normal glucose metabolism, DM2: type 2 diabetes mellitus, Fat: fat-rich meal, Carb: carbohydrate-rich meal, MPO: myeloperoxidase b P < 0.05 versus baseline


87

0 2 4 6 8

30

40

50

60

70

time [hours]

MPO [µµ µµg/L]

0 2 4 6 8

30

40

50

60

70

time [hours]

MPO [µµ µµg/L]

0 2 4 6 8

30

40

50

60

70

time [hours]

MPO [µµ µµg/L]

0 2 4 6 8

30

40

50

60

70

time [hours]

MPO [µµ µµg/L]

Figure 5.2. Time course of MPO (mean ± SEM) following two consecutive carbohydrate-rich

meals (solid line) [left panel] and two fat-rich meals (dotted line) [right panel], given at 0

and 4 hours in individuals with normal glucose metabolism (NGM: open dots) and patients

with type diabetes mellitus (DM2: solid dots)

DISCUSSION

This is, to the best of our knowledge, the first study to examine postprandial MPO

responses. We found no significant differences in baseline MPO between DM2 and

NGM. Baseline MPO was negatively related to HDL-cholesterol and positively

correlated to leukocyte counts. Total leukocytes increased after the fat-rich meals

in both NGM and DM2, whereas MPO decreased postprandially in NGM (both meals)

and in DM2 (fat-rich meals).

Previous studies have shown that leukocytes increase following a fat-rich meal [9].

In addition, one study suggested that activation markers on monocytes and

neutrophils are increased and related to postprandial changes in triglycerides [8].

Furthermore, activation markers on monocytes and neutrophils were higher in DM2

as compared to healthy controls [23]. These observations suggest that postprandial

leukocyte recruitment and activation may be a mechanism in the relation between

postprandial derangements and increased CVD risk. In addition, a number of

studies suggest that patients with higher MPO levels have an increased CVD risk

[15-18]. In the present study, the number of leukocytes in the fasting blood

samples was higher in DM2 as compared to NGM, but in spite of this, MPO was not

significantly higher in DM2 as compared to NGM. This lack of difference may be

explained by an impaired leukocyte function in DM2. Indeed, a previous study

CHAPTER 5

88

showed lower MPO in lymphocytes and neutrophils in patients with DM2 compared

to healthy controls [24]. In contrast to our initial hypothesis, we found lower MPO

values in the postprandial state after both meal types. Therefore, changes may not

only be meal-induced, but also due to diurnal changes in leukocytes and MPO.

Future studies assessing diurnal MPO course in the fasting state should address

this issue. In addition, because of unavailable plasma samples at 6 hours

postprandial, the present data may have underestimated the true changes in

postprandial MPO.

Of interest, we found a negative correlation between MPO and HDL-cholesterol.

Indeed, MPO binds to apolipoprotein AI of HDL particles and may subsequently be

inactivated or excreted [25]. Furthermore, MPO may be involved in generating

dysfunctional HDL and thus promote atherogenesis by this pathway [26]. In line

with this conjecture, a recent study has shown that MPO is associated with

progression of carotid artery stenosis in patients with low HDL-cholesterol levels

[27]. MPO has also been shown to promote oxidation of LDL particles [28,29],

which may provide another mechanism linking MPO to CVD risk.

We measured MPO in EDTA-plasma instead of heparin-plasma or serum. MPO mass

determined in serum is in general higher than in heparin plasma due to in vitro

release of MPO from activated leukocytes [30]. For heparin plasma it is necessary to

collect the samples on ice before sample processing (i.e. centrifugation) to avoid

MPO release [30]. When EDTA-blood was taken, we observed no changes in MPO

concentrations over a period of 2 hours at room temperature before sample

processing, whereas MPO rose gradually in heparin-plasma as well as in blood

withdrawn in tubes without an added anticoagulant (data not shown). Hence, EDTA-

plasma was chosen for MPO measurements because we assume that this may

reflect the in vivo MPO secretion of leukocytes most precisely. Although Zhang and

coworkers reported a high correlation between MPO mass and MPO activity (r =

0.95) [15], we cannot exclude that postprandial changes in MPO activity are more

pronounced than alterations in MPO concentrations.

In conclusion, we found no increase in MPO concentration despite the postprandial

increase in leukocytes after the fat-rich meals, suggesting that postprandial

leukocyte recruitment is likely to occur, but without a concomitant MPO release in

plasma. Our findings provide no support to the initial hypothesis that meal-induced

release of MPO might be a mechanism that contributes to CVD risk in


89

postmenopausal women with NGM or DM2 and therefore other mechanisms may

link postprandial leukocytes to CVD risk.

ACKNOWLEDGEMENTS

This study was financially supported by funding from Novartis AG Switzerland, and

by a grant from the Dutch Diabetes Research Foundation (grant no.2001.00.052).

The myeloperoxidase ELISA kits were generously provided by Mercodia AB,

Uppsala, Sweden. The authors thank Ina Kamphuis, Harry Twaalfhoven and Rick

Vermue for performing the MPO-assays and all participants for their contributions.

The authors have no conflict of interest with respect to this study.

CHAPTER 5

90

REFERENCES

1. Kannel WB, McGee DL. Diabetes and glucose tolerance as risk factors for

cardiovascular disease: the Framingham study. Diabetes Care 1979;2:120-6.

2. Stengard JH, Tuomilehto J, Pekkanen J, et al. Diabetes mellitus, impaired glucose

tolerance and mortality among elderly men: the Finnish cohorts of the Seven

Countries Study. Diabetologia 1992;35:760-5.

3. Haffner SM, Lehto S, Ronnemaa T, Pyorala K, Laakso M. Mortality from coronary heart

disease in subjects with type 2 diabetes and in nondiabetic subjects with and without

prior myocardial infarction. N Engl J Med 1998;339:229-34.

4. Lusis AJ. Atherosclerosis. Nature 2000;407:233-41.

5. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med 1999;340:115-26.

6. Friedman GD, Klatsky AL, Siegelaub AB. The leukocyte count as a predictor of


7. Phillips AN, Neaton JD, Cook DG, Grimm RH, Shaper AG. Leukocyte count and risk of

major coronary heart disease events. Am J Epidemiol 1992;136:59-70.

8. Van Oostrom AJ, Rabelink TJ, Verseyden C, et al. Activation of leukocytes by

postprandial lipemia in healthy volunteers. Atherosclerosis 2004;177:175-82.

9. Van Oostrom AJ, Sijmonsma TP, Verseyden C, et al. Postprandial recruitment of


10. Tushuizen ME, Diamant M, Heine RJ. Postprandial dysmetabolism and cardiovascular

disease in type 2 diabetes. Postgrad Med J 2005;81:1-6.



1999;19:2737-41.

12. Nauseef WM. Insights into myeloperoxidase biosynthesis from its inherited

deficiency. J Mol Med 1998;76:661-8.

13. Winterbourn CC. Comparative reactivities of various biological compounds with

myeloperoxidase-hydrogen peroxide-chloride, and similarity of the oxidant to

hypochlorite. Biochim Biophys Acta 1985;840:204-10.

14. Lau D, Baldus S. Myeloperoxidase and its contributory role in inflammatory vascular

disease. Pharmacol Ther 2006;111:16-26.

15. Zhang R, Brennan ML, Fu X, et al. Association between myeloperoxidase levels and

risk of coronary artery disease. JAMA 2001;286:2136-42.

16. Baldus S, Heeschen C, Meinertz T, et al. Myeloperoxidase serum levels predict risk in

patients with acute coronary syndromes. Circulation 2003;108:1440-5.

17. Tang WH, Brennan ML, Philip K, et al. Plasma myeloperoxidase levels in patients with

chronic heart failure. Am J Cardiol 2006;98:796-9.

18. Vita JA, Brennan ML, Gokce N, et al. Serum myeloperoxidase levels independently

predict endothelial dysfunction in humans. Circulation 2004;110:1134-9.


91

19. Zhou T, Zhou Sh, Qi Ss, et al. The effect of atorvastatin on serum myeloperoxidase

and CRP levels in patients with acute coronary syndrome. Clinica Chimica Acta

2006;368:168-72.

20. Kumar AP, Reynolds WF. Statins downregulate myeloperoxidase gene expression in

macrophages. Biochem Biophys Res Commun 2005;331:442-51.



1995;18:1270-3.




23. Van Oostrom AJ, van Wijk JP, Sijmonsma TP, Rabelink TJ, Castro CM. Increased

expression of activation markers on monocytes and neutrophils in type 2 diabetes.

Neth J Med 2004;62:320-5.

24. Uchimura K, Nagasaka A, Hayashi R, et al. Changes in superoxide dismutase activities

and concentrations and myeloperoxidase activities in leukocytes from patients with

diabetes mellitus. J Diabetes Complications 1999;13:264-70.

25. Zheng L, Nukuna B, Brennan ML, et al. Apolipoprotein A-I is a selective target for

myeloperoxidase-catalyzed oxidation and functional impairment in subjects with

cardiovascular disease. J Clin Invest 2004;114:529-41.

26. Nicholls SJ, Zheng L, Hazen SL. Formation of dysfunctional high-density lipoprotein by

myeloperoxidase. Trends Cardiovasc Med 2005;15:212-9.

27. Exner M, Minar E, Mlekusch W, et al. Myeloperoxidase predicts progression of carotid

stenosis in states of low high-density lipoprotein cholesterol. J Am Coll Cardiol

2006;47:2212-8.

28. Kostyuk VA, Kraemer T, Sies H, Schewe T. Myeloperoxidase/nitrite-mediated lipid

peroxidation of low-density lipoprotein as modulated by flavonoids. FEBS Lett

2003;537:146-50.

29. Steffen Y, Wiswedel I, Peter D, Schewe T, Sies H. Cytotoxicity of

myeloperoxidase/nitrite-oxidized low-density lipoprotein toward endothelial cells is

due to a high 7beta-hydroxycholesterol to 7-ketocholesterol ratio. Free Radic Biol Med

2006;41:1139-50.

30. Chang PY, Wu TL, Hung CC, et al. Development of an ELISA for myeloperoxidase on

microplate: normal reference values and effect of temperature on specimen

preparation. Clin Chim Acta 2006;373:158-63.

CHAPTER 5

92

93

Part 2

LIVER ENZYMES IN RELATION TO METABOLIC SYNDROME, TYPE 2 DIABETES MELLITUS AND CARDIOVASCULAR RISK

95

Chapter 6

ALANINE AMINOTRANSFERASE AS A MARKER OF NON-ALCOHOLIC FATTY LIVER DISEASE IN RELATION TO TYPE 2 DIABETES MELLITUS

AND CARDIOVASCULAR DISEASE


Michaela Diamant1


Maarten E. Tushuizen1

Tom Teerlink3

Robert J. Heine1,2

From the Departments of 1Endocrinology/Diabetes Center and 3Clinical Chemistry,

and the 2EMGO Institute, VU University Medical Center, Amsterdam, The

Netherlands

Diabetes Metabolism Research and Reviews 2006;22:437-443

Copyright © 2006 John Wiley & Sons Ltd.

CHAPTER 6

96

ABSTRACT

For a long time, hepatic steatosis was believed to be a benign condition. Only

recently, liver steatosis, also termed non-alcoholic fatty liver disease (NAFLD), has

gained much interest. In most cases of NAFLD, a condition regardedas the hepatic

component of the metabolic syndrome, the enzyme alanine aminotransferase (ALT)

is elevated and consequently has been used as a marker for NAFLD. More recently,

several cross-sectional and prospective studies have demonstrated associations of

this liver enzyme with features of the metabolic syndrome and type 2 diabetes

mellitus. This review discusses the biochemical and metabolic properties of ALT, its

applicability as a marker of NAFLD and describes its possible role in the

pathogenesis of the metabolic syndrome and type 2 diabetes mellitus and

subsequent cardiovascular disease. In addition treatment strategies to ameliorate

NAFLD and the associated risks are discussed.

ALANINE AMINOTRANSFERASE, TYPE 2 DIABETES MELLITUS AND CARDIOVASSCULAR DISEASE

97

INTRODUCTION

For a long time, hepatic steatosis was believed to be a benign condition. Only

recently, liver steatosis, also termed non-alcoholic fatty liver disease (NAFLD), has

gained much interest. NAFLD, which is characterized by a wide spectrum of liver

pathology ranging from mere liver steatosis to the more severe non-alcoholic

steatohepatitis, resembles alcohol-induced liver disease, but develops in subjects

who are not heavy alcohol consumers and have negative tests for viral and auto-

immune liver diseases [1–3]. This condition was first described in detail by Ludwig

and colleagues in 1980, who reported 20 moderately obese patients with liver

biopsy changes resembling alcohol-induced hepatitis, although none of these

patients reported a history of alcohol abuse [4]. Only in recent years, NAFLD has

gained appreciation as a pathogenic factor of insulin resistance and type 2 diabetes

mellitus. Furthermore, several studies showed an association between NAFLD and

features of the metabolic syndrome, including dyslipidemia and (visceral) obesity,

stressing the association with insulin resistance as an important feature of NAFLD.

Currently, NAFLD is considered by some authors to be the hepatic component of

the metabolic syndrome [5,6].

In the majority of cases, NAFLD causes asymptomatic elevation of the level of liver

enzymes alanine aminotransferase (ALT), aspartate aminotransferase (AST) and γ-

glutamyltransferase (GGT)) [7]. Of these liver enzymes, ALT is most closely related

to liver fat accumulation [8], and consequently ALT has been used as marker of

NAFLD. GGT may also be elevated in patients with NAFLD, but only limited data on

its occurrence and degree of elevation are available. Furthermore, the correlation of

liver fat measured by proton-MR spectroscopy with GGT is markedly lower than

with ALT [8], possibly because GGT is also produced in other tissues [9,10]. Finally,

to the best of our knowledge no direct validation of GGT with fat content in liver

biopsy specimen has been reported. Several cross-sectional studies have found

associations of ALT with several features of the metabolic syndrome. At present,

ALT is often used in epidemiological studies as a surrogate marker for NAFLD. In

addition, recent prospective epidemiological studies have demonstrated that ALT is

associated with future risk of type 2 diabetes mellitus and the metabolic syndrome

[11–17].

This review discusses the biochemical and metabolic properties of ALT, its

applicability as a marker of NAFLD and describes its possible role in the

pathogenesis of the metabolic syndrome and type 2 diabetes mellitus and

CHAPTER 6

98

subsequent cardiovascular disease (CVD), as well as discusses the treatment

strategies to ameliorate NAFLD and the associated risks.

BIOCHEMISTRY AND LABORATORY ANALYSIS

ALT catalyses the reversible transamination between L-alanine and α-ketoglutarate

to form pyruvate and L-glutamate and as such has an important role in

gluconeogenesis and amino-acid metabolism (Figure 6.1). The reaction is

reversible, but the equilibrium of the ALT reaction favors the formation of L-

alanine. The vitamin B6 vitamers pyridoxal-5’-phosphate (PLP) and its amino

analogue pyridoxamine-5’-phosphate, function as coenzymes in the amino transfer

reaction. ALT enzyme activity is primarily found in the liver, but its activity,

although much lower, is also found in many other organs including muscle, heart,

kidney, brain and adipose tissue [18-20].

Figure 6.1. Alanine aminotransferase (ALT) catalyses the interconversion of the amino-group from L-

alanine to α-ketoglutarate and forms pyruvate and L-glutamate. Pyridoxal-5’-phosphate (PLP)

functions as a coenzyme in the amino-transfer reaction.

In 1957, Wroblewski and Cabaud described a colorimetric ALT assay based on

measurement of glutamate by paper chromatography [21]. In later assays, the

transamination reaction was coupled to a second reaction involving the reduction

of pyruvate to lactate by lactate dehydrogenase [22]. In this reaction nicotinamide

COO-

|

H _C _ NH2 +

|

CH3

COO-

|

C = O

|

CH2 |

CH2 |

COO-

COO-

|

C = O +

|

CH3

COO-

|

H _C _ NH2

|

CH2 |

CH2 |

COO-

ALT, PLP

L-alanine αααα-ketoglutarate pyruvate L-glutamate

COO-

|

H _C _ NH2 +

|

CH3

COO-

|

C = O

|

CH2 |

CH2 |

COO-

COO-

|

C = O +

|

CH3

COO-

|

H _C _ NH2

|

CH2 |

CH2 |

COO-

ALT, PLP

L-alanine αααα-ketoglutarate pyruvate L-glutamate


99

adenine dinucleotide (NADH) is oxidized to NAD+ and the decrease in NADH is

assessed by measuring the absorption at 340 or 334 nm. In the past, comparison

between ALT activities reported by different laboratories was hampered by the fact

that assay temperatures ranging from 25 to 37 ˚C were used, leading to large

inter-laboratory differences in reported ALT activities and reference values. The

International Federation of Clinical Chemistry and Laboratory Medicine established

in 2002 a reference system for the measurement of enzyme activity of clinically

important enzymes, including ALT, to be measured at 37 ˚C [23]. Levels of 10 to

45 U/l are considered as normal although reference values may still vary among

laboratories [24]. ALT activity is stable for 24 hours in whole blood, with a marked

decrease thereafter. In serum, ALT enzyme activity is stable for three days at room

temperature and for three weeks in the refrigerator, but a marked decrease is

observed after repeated freezing and thawing [25-28]. It is therefore strongly

advisable not to determine ALT activity in frozen samples, as a falsely decreased

ALT activity may be found. Day-to-day variations of 10-30% have been observed

[29]. Moreover, significant diurnal variations have been reported in both healthy

subjects and in patients with chronic liver disease, necessitating standardized

blood sampling times when comparing enzyme values in clinical practice and

research [30].

Early biochemical studies have suggested the existence of two isoforms of ALT in

humans. Based on the amino acid sequence of liver ALT [31], Sohocki and

colleagues have cloned the human ALT1 located on human chromosome 8q24.3

[32]. Recently, Yang and colleagues cloned the second isoform ALT2 that was

mapped to the human chromosome 16q12.1 and demonstrated that this isoform

was mainly expressed in muscle and adipose tissue [33]. The expression of ALT2 in

adipose tissue may indicate that this isoform may contribute to the homeostasis of

fatty acid metabolism and storage. In fatty livers of obese mice, Jadhao and

colleagues demonstrated that murine ALT2 gene expression was induced 2-fold

whereas expression of murine ALT1 remained the same, and the total hepatic ALT

enzyme activity was elevated by 30% [34]. These observations indicate that ALT2

may be responsible for the increased ALT activity in hepatic steatosis. In the

pathogenesis of NAFLD, the role of different isoforms in humans has not been

studied yet. With the standard ALT assay used in clinical practice, enzyme activity

of both ALT isoforms is measured.

CHAPTER 6

100

ALT AND FATTY LIVER

Fatty liver content can be assessed with different methods. Direct measurement of

hepatic fat by means of liver biopsies, which is regarded as the “gold standard”

[35], although questionable because of the unknown distribution of fatty

infiltration in the liver, has only limited applicability in epidemiological and clinical

studies, because of the elaborate and invasive nature and associated risks of this

technique. Furthermore, no consensus exists on the standardization of diagnostic

criteria bases on liver histology, although only recently a scoring system for NAFLD

has been proposed [36]. Imaging techniques, such as ultrasound, computed

tomography and proton magnetic resonance spectroscopy (1H-MRS), are non-

invasive and can be used rather than biopsies [37,38]. Ultrasound of the liver is

another non-invasive technique which can be used to indicate the presence or

absence of fatty infiltration of the liver, however, at present, this technique at its

best can provide only semi-quantitative data with regard to NAFLD. However, non-

invasive techniques cannot differentiate between fatty liver alone or steatohepatitis. 1H-MRS has been validated against direct determination of triglyceride content in

human liver biopsies [39], and may be used for clinical purposes, but has limited

applicability in larger populations, because of logistics and high costs. In more

recent epidemiological studies, ALT has been used as a surrogate marker for liver

fat accumulation. This was first described by Nanji and colleagues in 1986, who

reported an association of liver enzymes (i.e. the ALT to AST ratio) with fatty liver in

obese patients [40]. Studies assessing the specificity and sensitivity of ALT as

marker of NAFLD are limited [41,42]. Prati and colleagues have suggested new cut-

off values for ALT in order to facilitate the identification of subjects with NAFLD

[43]. However, since cut-off values depend on the assay used, this proposal has

been questioned by others [24,44]. One study has assessed the correlation of ALT

with 1H-MRS and found a modest, but significant correlation (r=0.5) [8]. We found a

similar correlation between liver fat accumulation measured by 1H-MRS and ALT

(r=0.58; P<0.001) in 36 healthy men and men with the metabolic syndrome with

and without hyperglycemia (type 2 diabetes mellitus) (M.E. Tushuizen et al.,

unpublished data).

In order to avoid confounding by other factors that might cause ALT elevation, a

detailed history on alcohol intake is mandatory to exclude those patients with ALT

elevation caused by alcohol-abuse, and to address alcohol-intake as a possible

confounder or effect-modifier. Viral and auto-immune liver disease and hepatotoxic

medication should be considered as well. With these limitations in mind, ALT is an


101

acceptable marker for hepatic steatosis in epidemiological studies; its use for

diagnostic and monitoring purposes in clinical care needs to be further delineated.

ALT AND INCIDENT METABOLIC SYNDROME AND TYPE 2 DIABETES

MELLITUS

The putative role of the liver in the pathogenesis of type 2 diabetes mellitus has

gained much interest and several cross-sectional and prospective studies have

shown associations of ALT with type 2 diabetes mellitus and the metabolic

syndrome. Several cross-sectional studies have demonstrated that ALT is related to

features of the metabolic syndrome and type 2 diabetes mellitus [7,45-48]. Clark

and colleagues for example found in their analysis of The National Health and

Nutrition Examination Survey (NHANES) that up to 31% of the elevated

aminotransferase activity could be explained by high alcohol consumption,

hepatitis B or C infection and/or high transferrin saturation. In the remaining 69%,

the elevated ALT activity was significantly associated with higher body mass index,

waist circumference, triglycerides, fasting insulin and lower HDL-cholesterol.

Additionally, in women, this elevation was associated with type 2 diabetes mellitus

and hypertension [7]. In one study it was demonstrated that ALT can be normal in

uncomplicated obesity [49], but this does not guarantee absence of NAFLD, as

normal ALT levels have been observed in patients with NAFLD, diagnosed by liver

biopsy, which is regarded as the gold standard in the diagnosis of NAFLD [50].

Several studies addressed the prospective relation of ALT with the metabolic

syndrome and type 2 diabetes mellitus (Table 6.1). Nakanishi and co-workers found

that of the liver enzymes studied, ALT was associated with future risk of metabolic

syndrome in middle-aged Japanese men, but used body mass index instead of

waist circumference to define the metabolic syndrome [16]. Hanley and co-workers

studied the relation of four different liver enzymes (including ALT) with the

development of the metabolic syndrome in a multiethnic cohort and demonstrated

that ALT was associated with an increased risk of incident metabolic syndrome

[17]. In patients with type 2 diabetes mellitus, elevated serum ALT enzyme activity

is more frequently observed than in the general population [54,55]. Some [11-15],

but not all studies [16,51-53] have demonstrated independent and significant

associations of ALT with incident type 2 diabetes mellitus. In 1988, Ohlson and

colleagues demonstrated that baseline ALT was a predictor of incident type 2

diabetes mellitus after 13.5 years of follow-up in a cohort of 766 Swedish men,

with a significant four-fold risk for those men in the upper quintile compared to

CHAPTER 6

102

those in the lower quintile. In the final multivariate model, ALT was a predictor of

incident type 2 diabetes mellitus in 451 Pima Indians after an average of 6.9 years,

after adjustment for fasting blood glucose levels, body mass index, bilirubin,

systolic blood pressure, uric acid and a family history of diabetes [11]. Vozarova

and colleagues found that higher ALT (upper compared to lower decile) was a

significant and independent predictor of incident type 2 diabetes mellitus with a

two-fold increase risk, after adjustment for age, sex, body fat, insulin sensitivity

and acute insulin response [12]. Other studies, mainly performed in men,

confirmed these earlier reports [13,14], while recent population-based studies

could not demonstrate independent associations of ALT with risk of type 2 diabetes

mellitus [52,53]. The observed association between ALT and incident type 2

diabetes mellitus in these reports may thus be explained by the fact that these

studies were performed in high-risk populations and may not be representative for

the general population. An alternative explanation is that the applied models over-

adjust for potential mediating variables, such as insulin and 2h glucose, that might

be directly related to liver fat and/or for risk factors clustered in the metabolic

syndrome. More population-based studies and a better understanding of the

pathophysiology are needed to confirm and understand the predictive value of ALT

in relation to the metabolic syndrome and type 2 diabetes mellitus.

ALT AND CARDIOVASCULAR DISEASE

Several studies have demonstrated associations between NAFLD or ALT and

markers of atherosclerosis or inflammation. Previously, we reported an association

of ALT with endothelial dysfunction, measured as flow-mediated vasodilatation

[56], and Kerner and colleagues found a relation of ALT with C-reactive protein

[57]. Only a limited number of studies have addressed the prospective relation of

ALT with cardiovascular disease and mortality. Arndt and colleagues studied the

association of ALT with all-cause mortality in 8,043 male construction workers and

found no significant association [58]. A recent study by Nakamura and co-workers

found a positive association of ALT with all-cause mortality in Japanese men and

women, but only for those with a body mass index below the median (22.7 kg/m2)

[59]. In the Hoorn Study, a population-based cohort of Caucasian men and women

aged 50 to 75 years, we studied the association of ALT at baseline with all-cause

mortality, incident cardiovascular and coronary heart disease events and found a

significant association of ALT with coronary heart disease after adjustment for the

other components of the metabolic syndrome and traditional risk factors [60].

ALANINE AMINOTRANSFERASE, T

YPE 2 DIABETES M

ELLITUS AND CARDIOVASSCULAR DISEASE

103

Table 5.1. Prospective studies assessing the relation of ALT and incident metabolic syndrome and type 2 diabetes mellitus

First Author Cohort Subjects Age Sex (%male) Follow -up, y Relative Risk Comparison Groups

Metabolic Syndrome

Nakanishi[16] Japanese men 2957 35-59 100% 7.0 1.42 (0.95 -2.11) Upper versus lower quintile

Hanley[17] IRAS 633 40-69 44% 5.2 2.12 (1.10 -4.09) Upper versus lower quartile

Type 2 diabetes mellitus

Ohlson[11] Swedish men 54 100% 13.5 3.9 (1.4 -11.1) Upper versus lower quintile

Vozar ova[12] Pima Indians 451 18-50 61% 6.9 1.9 (1.1 -3.3) Upper versus lower decile

Sattar[13] WOSCOPS 5974 45-64 100% 5.6 2.04 (1.16 -3.58) Upper versus lower quartile

Hanley[14] IRAS 906 40-69 44% 5.2 2.00 (1.22 -3.38) Upper versus lower three quartiles

Wann amethee[15] British Regional Heart

Study

3500 40-59 100% 5.0 1.91 (1.01 -3.60) Upper versus lower quartile


Nannipieri[51] Mexico City Diabetes

Study

1233 35-64 39% 7.0 No significant association Upper quartile vs. lower quartiles

Schindhelm[52] Hoorn Study 1289 50-75 46% 6.4 1.18 (0.66 –2.11) Upper tertile versus lower tertile

Andre[53] DESIR Study 2071

2130

30-65 100%

0%

3.0 1.5 (0.6 -30)

0.7 (0.2 -2.1)

Upper ver sus lower quartile



Metabolic Syndrome


Hanley[17] IRAS



Metabolic Syndrome






Sattar[13] WOSCOPS








Study


Nakanishi[16] Japanese men 3620 35-




Study




Study




2130

30-65 100%

59 100% 7.0 2.07 (1.07 -3.98) Upper versus lower quintile


Study




2130

30-65 100%

0%

3.0 1.5 (0.6 -30)

0.7 (0.2 -2.1)




Metabolic Syndrome


Hanley[17] IRAS



Metabolic Syndrome






Sattar[13] WOSCOPS








Study






Study




Study




2130

30-65 100%



Study




2130

30-65 100%

0%

3.0 1.5 (0.6 -30)

0.7 (0.2 -2.1)




Metabolic Syndrome


Hanley[17] IRAS



Metabolic Syndrome






Sattar[13] WOSCOPS








Study






Study




Study




2130

30-65 100%



Study




2130

30-65 100%

0%

3.0 1.5 (0.6 -30)

0.7 (0.2 -2.1)






Sattar[13] WOSCOPS








Study






Study




Study




2130

30-65 100%



Study




2130

30-65 100%

0%

3.0 1.5 (0.6 -30)

0.7 (0.2 -2.1)





Study






Study




Study




2130

30-65 100%



Study




2130

30-65 100%

0%

3.0 1.5 (0.6 -30)

0.7 (0.2 -2.1)


CHAPTER 6

104

PATHOGENIC MECHANISMS

To date, an elevated ALT is considered a consequence of hepatocyte damage due

to NAFLD. However, the measured plasma elevations of ALT may also be a

consequence of high systemic ALT2 isoform levels that is largely derived from

adipose tissue in obesity and insulin resistance, as has been observed in mice [34].

Insulin resistance, increased pro-inflammatory cytokine production, oxidative stress

and mitochondrial dysfunction leading to hepatocyte damage/destruction, all have

been posed as important pathophysiological mechanisms [61]. Indeed a recent

study reported on the association of ALT and GGT with markers of inflammation

and oxidative stress [62].

Adiponectin, released by adipose tissue has been implied in the pathogenesis of

NAFLD. Obese subjects and patients with type 2 diabetes mellitus have lower levels

of adiponectin compared to healthy controls. Yokoyama and colleagues found an

inverse association of ALT, AST and GGT with adiponectin levels in Japanese male

workers. These associations were independent of age, body mass index, insulin

resistance, serum triglycerides and total cholesterol [63]. Aygun and co-workers

found that adiponectin levels were lower in patients with biopsy proven NAFLD and

elevated liver enzymes, compared to patients with NAFLD and normal liver enzymes

and to those without NAFLD [64]. The exact mechanisms, by which adiponectin is

related to NAFLD are not fully elucidated and need further clarification.

Insulin resistance and visceral obesity have a major impact on regulatory processes

of the (postprandial) lipoprotein and glucose metabolism. The normal suppression

of lipolysis by insulin is impaired in the insulin resistant state and this leads to an

elevated release of free fatty acids from the adipose tissue and subsequent

accumulation in hepatocytes, further enhancing hepatic steatosis and contributing

to increased risk of atherothrombotic disease in subjects with NAFLD and insulin

resistance. Dietary fat composition also seems to influence the degree of hepatic

steatosis [65], although the majority of triglycerides arise from free fatty acids,

derived from adipose tissue lipolysis [66]. It has been demonstrated that a fatty

liver is insulin resistant, resulting in an elevated glucose and very low-density

lipoprotein production [12]. Furthermore, cytokines released from the (visceral)

adipose tissue, including interleukin-6 and tumor necrosis factor-alpha, both

associated with decreased hepatic insulin sensitivity [67], may further enhance fatty

infiltration and decrease hepatocyte integrity.

ALANINE AMINOTRANSFERASE, TYPE 2 DIABETES MELLITUS AND CARDIOVASCULAR DISEASE

105

Another explanation might be up-regulation of ALT enzyme activity. Among the

amino acids, alanine is the most effective precursor for gluconeogenesis [68].

Gluconeogenesis is increased in subjects with type 2 diabetes mellitus [69], due to

increased substrate delivery (e.g. alanine), and an increased conversion of alanine

to glucose [70]. This increased conversion might contribute to the increased ALT

activity, as previously observed in alloxan-diabetic rats [71].

ALT might thus be up-regulated as a compensatory response to the impaired

hepatic insulin signaling or, alternatively, may leak more easily out of the

hepatocytes as a consequence of fatty infiltrations and subsequent damage.

THERAPEUTIC INTERVENTIONS

Therapeutic approaches, including diets and exercise, leading to weight loss and

improvement of the cardiovascular disease risk profile, have also been shown to

lower liver fat content.. These interventions concomitantly improve insulin

sensitivity and/or reduce oxidative stress. Ueno and colleagues compared the

effect of restricted diet and exercise in a group of 15 obese subjects during 3

months to that of no lifestyle changes in a control group of 10 obese subjects.

They found a decrease in biochemical markers of NAFLD including ALT, total

cholesterol and fasting blood glucose. This was confirmed by liver biopsies

demonstrating significant lower fat in the treatment group compared to the control

group [72]. Similar results were found in a recent study by Suzuki and co-workers,

who studied the effect of diet and exercise induced weight reduction on the change

of serum ALT and found that subjects with more than 5% weight reduction and

persistence of this weight reduction showed improvement of serum ALT [73].

Hickman and colleagues assessed the effect of weight loss in patients with chronic

liver disease. In addition to improvements in quality of life and insulin sensitivity,

they also demonstrated that ALT reduction was correlated with the amount of

weight loss (r=0.35; P=0.04) after 15 months. ALT remained significantly lower in

those patients who maintained their reduced weight [74]. Although these life-style

interventions seem promising in the short term, long term effects have not yet

been established and might not be sufficient in patients with multiple

cardiovascular disease risk factors, such as subjects with the metabolic syndrome

and/or type 2 diabetes mellitus.

Several pharmacological agents, including metformin and peroxisome proliferator-

activated receptor-γ (PPAR-γ) agonists have been used in the treatment of NAFLD.

Only a limited number of studies addressed the effect of metformin and/or PPAR-γ

CHAPTER 6

106

agonists on ALT. An open-labeled trial of 20 mg/kg metformin for one year,

demonstrated only a transient improvement of liver enzymes [75]. Tiikkainen and

co-workers compared the effect of 2 g of metformin to 8 mg of the PPAR-γ agonists

rosiglitazone on liver fat content and hepatic insulin sensitivity in a 16 week

double-blind randomized trial in 20 subjects. They found that rosiglitazone

decreased liver fat by 51%, whereas no significant decrease was observed in the

metformin treated subjects [76]. In contrast, Bugianesi and co-workers found that 2

g of metformin lead to more ALT normalization (i.e. below the reference range)

than either vitamin E or diet induced weight reduction in 55 subjects in an 12

months open-labeled randomized trail. Metformin induced a decrease of 50% in

liver fat and a decrease of inflammation and necrosis, using paired liver biopsies

[77]. Unfortunately, in this study no paired biopsies were performed in the subjects

who were randomized to vitamin E or the weight reduction intervention, thus the

level of liver histology improvement in the treatment group could not be compared

with the effect in the control group. Sanyal and co-workers compared the effect of

vitamin E versus vitamin E and the PPAR-γ agonists pioglitazone (30 mg/day) in a

randomized clinical trial with 10 patients with non-alcoholic steatohepatitis in each

treatment arm. They found that vitamin E reduced the amount of hepatic steatosis,

whereas the combination of vitamin E with pioglitazone also improved liver

histology [78]. In a recently published study, Ding and co-workers demonstrated

that 60 days administration of a glucagon-like peptide-1 (GLP-1) receptor agonists,

an incretin-mimetic known to promote insulin production and secretion, reduced

hepatic steatosis in ob/ob mice [79]. Studies in humans assessing the effect of GLP-

1 receptor agonists on hepatic steatosis are needed to confirm these results and

assess the applicability of these agents for the treatment of hepatic steatosis. Long-

term effects and reduction of morbidity and mortality of the above mentioned

treatment modalities need further evaluation to achieve optimal treatment

strategies.

CONCLUSIONS

In epidemiological studies, ALT seems an appropriate marker of fatty liver disease,

if known factors that cause ALT elevation have been addressed properly (e.g

alcohol consumption and hepatitis B and C infections). ALT enzyme activity and

NAFLD are associated with cardiovascular risk factors and treatment of NAFLD

should be aimed at reducing hepatic fat content and reduce the concomitant CVD

risk. Moreover, ALT might provide guidance in clinical practice for the choice

between the different treatment options and to monitor the effect of therapy.


107

REFERENCES


2. Matteoni CA, Younossi ZM, Gramlich T, Boparai N, Liu YC, McCullough AJ.

Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity.

Gastroenterology 1999;116:1413-9.


2001;21:3-16.

4. Ludwig J, Viggiano TR, McGill DB, Oh BJ. Nonalcoholic steatohepatitis: Mayo Clinic

experiences with a hitherto unnamed disease. Mayo Clin Proc 1980;55:434-8.



6. Marchesini G, Brizi M, Bianchi G, et al. Nonalcoholic fatty liver disease: a feature of







2004;47:1360-1369.

9. Paolicchi A, Emdin M, Ghliozeni E, et al. Images in cardiovascular medicine. Human

atherosclerotic plaques contain gamma-glutamyl transpeptidase enzyme activity.

Circulation 2004;109:1440.

10. Pompella A, Emdin M, Passino C, Paolicchi A. The significance of serum gamma-

glutamyltransferase in cardiovascular diseases. Clin Chem Lab Med 2004;42:1085-91.

11. Ohlson LO, Larsson B, Bjorntorp P, et al. Risk factors for type 2 (non-insulin-

dependent) diabetes mellitus. Thirteen and one-half years of follow-up of the

participants in a study of Swedish men born in 1913. Diabetologia 1988;31:798-805.

12. Vozarova B, Stefan N, Lindsay RS, et al. High alanine aminotransferase is associated

with decreased hepatic insulin sensitivity and predicts the development of type 2

diabetes. Diabetes 2002;51:1889-95.

13. Sattar N, Scherbakova O, Ford I, et al. Elevated alanine aminotransferase predicts

new-onset type 2 diabetes independently of classical risk factors, metabolic

syndrome, and C-reactive protein in the west of Scotland coronary prevention study.

Diabetes 2004;53:2855-60.

14. Hanley AJ, Williams K, Festa A, et al. Elevations in markers of liver injury and risk of

type 2 diabetes: the insulin resistance atherosclerosis study. Diabetes 2004;53:2623-

32.

15. Wannamethee SG, Shaper AG, Lennon L, Whincup PH. Hepatic enzymes, the metabolic

syndrome, and the risk of type 2 diabetes in older men. Diabetes Care 2005;28:2913-

8.

CHAPTER 6

108

16. Nakanishi N, Suzuki K, Tatara K. Serum gamma-glutamyltransferase and risk of

metabolic syndrome and type 2 diabetes in middle-aged Japanese men. Diabetes Care

2004;27:1427-32.

17. Hanley AJ, Williams K, Festa A, et al. Liver markers and development of the metabolic

syndrome: the insulin resistance atherosclerosis study. Diabetes 2005;54:3140-7.

18. Awapara J, Seale B. Distribution of transaminases in rat organs. J Biol Chem

1952;194:497-502.

19. DeRosa G, Swick RW. Metabolic implications of the distribution of the alanine

aminotransferase isoenzymes. J Biol Chem 1975;250:7961-7.

20. Bhattacharya SB, Datta AG. Is brain a gluconeogenic organ? Mol Cell Biochem

1993;125:51-7.

21. Wroblewski F, Cabaud P. Colorimetric measurement of serum glutamic pyruvic

transaminase. Tech Bull Regist Med Technol 1957;27:13-7.

22. Bergmeyer HU, Scheibe P, Wahlefeld AW. Optimization of methods for aspartate

aminotransferase and alanine aminotransferase. Clin Chem 1978;24:58-73.

23. Schumann G, Bonora R, Ceriotti F, et al. IFCC primary reference procedures for the

measurement of catalytic activity concentrations of enzymes at 37 degrees C.

International Federation of Clinical Chemistry and Laboratory Medicine. Part 4.

Reference procedure for the measurement of catalytic concentration of alanine

aminotransferase. Clin Chem Lab Med 2002;40:718-24.

24. Kaplan MM. Alanine aminotransferase levels: what's normal? Ann Intern Med

2002;137:49-51.

25. Ono T, Kitaguchi K, Takehara M, Shiiba M, Hayami K. Serum-constituents analyses:

effect of duration and temperature of storage of clotted blood. Clin Chem

1981;27:35-38.

26. Williams KM, Williams AE, Kline LM, Dodd RY. Stability of serum alanine

aminotransferase activity. Transfusion 1987;27:431-3.

27. DiMagno EP, Corle D, O'Brien JF, et al. Effect of long-term freezer storage, thawing,

and refreezing on selected constituents of serum. Mayo Clin Proc 1989;64:1226-34.

28. Clark S, Youngman LD, Palmer A, et al. Stability of plasma analytes after delayed

separation of whole blood: implications for epidemiological studies. Int J Epidemiol

2003;32:125-30.

29. Fraser CG, Cummings ST, Wilkinson SP, et al. Biological variability of 26 clinical

chemistry analytes in elderly people. Clin Chem 1989;35:783-6.

30. Cordoba J, O'Riordan K, Dupuis J, Borensztajin J, Blei AT. Diurnal variation of serum

alanine transaminase activity in chronic liver disease. Hepatology 1998;28:1724-5.

31. Ishiguro M, Takio K, Suzuki M, et al. Complete amino acid sequence of human liver

cytosolic alanine aminotransferase (GPT) determined by a combination of

conventional and mass spectral methods. Biochemistry 1991;30:10451-7.


109

32. Sohocki MM, Sullivan LS, Harrison WR, et al. Human glutamate pyruvate transaminase

(GPT): localization to 8q24.3, cDNA and genomic sequences, and polymorphic sites.

Genomics 1997;40:247-52.

33. Yang RZ, Blaileanu G, Hansen BC, Shuldiner AR, Gong DW. cDNA cloning, genomic

structure, chromosomal mapping, and functional expression of a novel human

alanine aminotransferase. Genomics 2002;79:445-50.

34. Jadhao SB, Yang RZ, Lin Q, et al. Murine alanine aminotransferase: cDNA cloning,

functional expression, and differential gene regulation in mouse fatty liver.

Hepatology 2004;39:1297-1302.

35. Brunt EM, Neuschwander-Tetri BA, Oliver D, Wehmeier KR, Bacon BR. Nonalcoholic

steatohepatitis: histologic features and clinical correlations with 30 blinded biopsy

specimens. Hum Pathol 2004;35:1070-82.

36. Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological

scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41:1313-21.

37. Saverymuttu SH, Joseph AE, Maxwell JD. Ultrasound scanning in the detection of

hepatic fibrosis and steatosis. Br Med J (Clin Res Ed) 1986;292:13-5.

38. Ricci C, Longo R, Gioulis E, et al. Noninvasive in vivo quantitative assessment of fat

content in human liver. J Hepatol 1997;27:108-13.

39. Szczepaniak LS, Babcock EE, Schick F, et al. Measurement of intracellular triglyceride

stores by H spectroscopy: validation in vivo. Am J Physiol 1999;276:E977-89.

40. Nanji AA, French SW, Freeman JB. Serum alanine aminotransferase to aspartate

aminotransferase ratio and degree of fatty liver in morbidly obese patients. Enzyme

1986;36:266-9.

41. Nomura K, Yano E, Shinozaki T, Tagawa K. Efficacy and effectiveness of liver

screening program to detect fatty liver in the periodic health check-ups. J Occup

Health 2004;46:423-8.

42. Kunde SS, Lazenby AJ, Clements RH, Abrams GA. Spectrum of NAFLD and diagnostic

implications of the proposed new normal range for serum ALT in obese women.

Hepatology 2005;42:650-6.

43. Prati D, Taioli E, Zanella A, et al. Updated definitions of healthy ranges for serum

alanine aminotransferase levels. Ann Intern Med 2002;137:1-10.

44. Ratziu V, Imbert-Bismut F, Messous D, Poynard T. The elusiveness of "normal" ALT in

fatty liver. Hepatology 2004;39:1172.

45. Ruhl CE, Everhart JE. Determinants of the association of overweight with elevated

serum alanine aminotransferase activity in the United States. Gastroenterology

2003;124:71-9.

46. Ioannou GN, Weiss NS, Boyko EJ, Kahn SE, Lee SP. Contribution of metabolic factors to

alanine aminotransferase activity in persons with other causes of liver disease.


47. Friedman LS, Dienstag JL, Watkins E, et al. Evaluation of blood donors with elevated

serum alanine aminotransferase levels. Ann Intern Med 1987;107:137-44.

CHAPTER 6

110

48. Wejstal R, Hansson G, Lindholm A, Norkrans G. Persistent alanine aminotransferase

elevation in healthy Swedish blood donors--mainly caused by obesity. Vox Sang

1988;55:152-6.

49. Iacobellis G, Moschetta A, Ribaudo MC, et al. Normal serum alanine aminotransferase

activity in uncomplicated obesity. World J Gastroenterol 2005;11:6018-21.

50. Mofrad P, Contos MJ, Haque M, et al. Clinical and histologic spectrum of nonalcoholic

fatty liver disease associated with normal ALT values. Hepatology 2003;37:1286-92.

51. Nannipieri M, Gonzales C, Baldi S, et al. Liver enzymes, the metabolic syndrome, and

incident diabetes: the Mexico City diabetes study. Diabetes Care 2005;28:1757-62.

52. Schindhelm RK, Dekker JM, Nijpels G, Heine RJ, Diamant M. No independent

association of alanine aminotransferase with risk of future type 2 diabetes in the

Hoorn study. Diabetes Care 2005;28:2812.

53. Andre P, Balkau B, Born C, et al. Hepatic markers and development of type 2 diabetes

in middle aged men and women: a three-year follow-up study. Diabetes Metab 2005;

31: 542-550.

54. Salmela PI, Sotaniemi EA, Niemi M, Maentausta O. Liver function tests in diabetic

patients. Diabetes Care 1984;7:248-54.

55. Meltzer AA, Everhart JE. Association between diabetes and elevated serum alanine

aminotransferase activity among Mexican Americans. Am J Epidemiol 1997;146:565-

71.

56. Schindhelm RK, Diamant M, Bakker SJL, et al. Liver alanine aminotransferase, insulin

resistance and endothelial dysfunction in normotriglyceridaemic subjects with type 2

diabetes mellitus. Eur J Clin Invest 2005;35:369-74.




58. Arndt V, Brenner H, Rothenbacher D, et al. Elevated liver enzyme activity in

construction workers: prevalence and impact on early retirement and all-cause

mortality. Int Arch Occup Environ Health 1998;71:405-12.

59. Nakamura K, Okamura T, Kanda H, et al. The Value of Combining Serum Alanine

Aminotransferase Levels and Body Mass Index to Predict Mortality and Medical Costs:

a 10-year Follow-up Study of National Health Insurance in Shiga, Japan. J Epidemiol

2006;16:15-20.

60. Schindhelm RK, Dekker JM, Nijpels G, et al. Alanine aminotransferase predicts

coronary heart disease events, a 10-year follow-up of the Hoorn study.


61. Day CP. Pathogenesis of steatohepatitis. Best Pract Res Clin Gastroenterol

2002;16:663-78.

62. Yamada J, Tomiyama H, Yambe M, et al. Elevated serum levels of alanine

aminotransferase and gamma glutamyltransferase are markers of inflammation and


111

oxidative stress independent of the metabolic syndrome. Atherosclerosis

2006;189:198-205.

63. Yokoyama H, Hirose H, Ohgo H, Saito I. Inverse association between serum

adiponectin level and transaminase activities in Japanese male workers. J Hepatol

2004;41:19-24.

64. Aygun C, Senturk O, Hulagu S, et al. Serum levels of hepatoprotective peptide

adiponectin in non-alcoholic fatty liver disease. Eur J Gastroenterol Hepatol

2006;18:175-80.

65. Westerbacka J, Lammi K, Hakkinen AM, et al. Dietary fat content modifies liver fat in

overweight nondiabetic subjects. J Clin Endocrinol Metab 2005;90:2804-9.

66. Donnelly KL, Smith CI, Schwarzenberg SJ, et al. Sources of fatty acids stored in liver

and secreted via lipoproteins in patients with nonalcoholic fatty liver disease. J Clin

Invest 2005;115:1343-51.

67. Kern PA, Ranganathan S, Li C, Wood L, Ranganathan G. Adipose tissue tumor necrosis

factor and interleukin-6 expression in human obesity and insulin resistance. Am J

Physiol Endocrinol Metab 2001;280:E745-51.

68. Yamamoto H, Aikawa T, Matsutaka H, Okuda T, Ishikawa E. Interorganal relationships

of amino acid metabolism in fed rats. Am J Physiol 1974;226:1428-33.

69. Zawadzki JK, Wolfe RR, Mott DM, et al. Increased rate of Cori cycle in obese subjects

with NIDDM and effect of weight reduction. Diabetes 1988;37:154-159.

70. Consoli A, Nurjhan N, Reilly JJ, Jr., Bier DM, Gerich JE. Mechanism of increased

gluconeogenesis in noninsulin-dependent diabetes mellitus. Role of alterations in

systemic, hepatic, and muscle lactate and alanine metabolism. J Clin Invest

1990;86:2038-45.

71. Rosen F, Roberts NR, Nichol CA. Glucocorticosteroids and transaminase activity. I.

Increased activity of glutamicpyruvic transaminase in four conditions associated with

gluconeogenesis. J Biol Chem 1959;234:476-80.

72. Ueno T, Sugawara H, Sujaku K, et al. Therapeutic effects of restricted diet and

exercise in obese patients with fatty liver. J Hepatol 1997;27:103-7.

73. Suzuki A, Lindor K, Saver JS, et al. Effect of changes on body weight and lifestyle in

nonalcoholic fatty liver disease. J Hepatol 2005;43:1060-6.

74. Hickman IJ, Jonsson JR, Prins JB, et al. Modest weight loss and physical activity in

overweight patients with chronic liver disease results in sustained improvements in

alanine aminotransferase, fasting insulin, and quality of life. Gut 2004;53:413-9.

75. Nair S, Diehl AM, Wiseman M, Farr GH, Jr., Perrillo RP. Metformin in the treatment of

non-alcoholic steatohepatitis: a pilot open label trial. Aliment Pharmacol Ther

2004;20:23-8.

76. Tiikkainen M, Hakkinen AM, Korsheninnikova E, et al. Effects of rosiglitazone and

metformin on liver fat content, hepatic insulin resistance, insulin clearance, and gene

expression in adipose tissue in patients with type 2 diabetes. Diabetes 2004;53:2169-

76.

CHAPTER 6

112

77. Bugianesi E, Gentilcore E, Manini R, et al. A randomized controlled trial of metformin

versus vitamin E or prescriptive diet in nonalcoholic fatty liver disease. Am J

Gastroenterol 2005;100:1082-90.

78. Sanyal AJ, Mofrad PS, Contos MJ, et al. A pilot study of vitamin E versus vitamin E and

pioglitazone for the treatment of nonalcoholic steatohepatitis. Clin Gastroenterol

Hepatol 2004;2:1107-15.

79. Ding X, Saxena NK, Lin S, Gupta N, Anania FA. Exendin-4, a glucagon-like protein-1

(GLP-1) receptor agonist, reverses hepatic steatosis in ob/ob mice. Hepatology

2006;43:173-81.

113

Chapter 7

LIVER ALANINE AMINOTRANSFERASE, INSULIN RESISTANCE AND ENDOTHELIAL DYSFUNCTION IN NORMOTRIGLYCERIDEMIC SUBJECTS

WITH TYPE 2 DIABETES MELLITUS


Michaela Diamant1

Stephan J. L. Bakker2

Rob A. J. M. van Dijk3

Peter G. Scheffer4

Tom Teerlink4

Piet J. Kostense5

Robert J. Heine1

From the Departments of 1Endocrinology/Diabetes Center, 3Radiology, 4Clinical

Chemistry and 5Clinical Epidemiology and Biostatistics, VU University Medical

Center, Amsterdam, The Netherlands and the 2Department of Internal Medicine,

University Hospital Groningen, Groningen, The Netherlands.

European Journal Clinical Investigation 2005;35:369-74

Copyright © 2005 Blackwell Publishing Ltd.

Reprinted with permission.

CHAPTER 7

114

ABSTRACT

Background Plasma levels of liver transaminases, including alanine

aminotransferase (ALT), are elevated in most cases of nonalcoholic fatty liver

disease (NAFLD). Elevated ALT levels are associated with insulin resistance, and

subjects with NAFLD have features of the metabolic syndrome that confer high-risk

cardiovascular disease. Alanine aminotransferase predicts the development of type

2 diabetes (DM2) in subjects with the metabolic syndrome. However, the role of

elevated ALT levels in subjects with overt DM2 has yet not been explored.

Materials and methods In a cross-sectional study, 64 normotriglyceridemic

subjects with DM2 were studied with regard to the relation between liver

transaminases with whole-body insulin sensitivity, measured with the euglycemic

hyperinsulinemic clamp and with brachial artery flow-mediated dilation (FMD) as a

marker of endothelial dysfunction.

Results On average, patients were normotriglyceridemic (plasma triglycerides

1.3±0.4 mmol/l) and had good glycemic control (HbA1c 6.2±0.8%). The mean ALT

level was 15.0±7.5 U/l, and the mean aspartate aminotransferase concentration

equaled 10.6±2.6 U/l. Alanine aminotransferase levels were negatively associated

with whole-body insulin sensitivity as well as with FMD (both P = 0.03, in

multivariate analyses; regression coefficients beta [95%CI]: -0.76 [-1.4 to –0.08] and

–0.31 [-0.58 to –0.03] respectively).

Conclusions In metabolically well-controlled patients with DM2, ALT levels are

related to decreased insulin-sensitivity and an impaired conduit vessel vascular

function.

ALANINE AMINOTRANSFERASE, INSULIN RESISTANCE AND ENDOTHELIAL DYSFUNCTION

115

INTRODUCTION

Non-alcoholic fatty liver disease (NAFLD) includes a wide spectrum of liver

pathology, ranging from fatty liver alone to the more severe non-alcoholic

steatohepatitis [1]. NAFLD develops in people who are not heavy alcohol

consumers, and usually has a benign clinical course [2,3]. Recent studies have

shown associations between NAFLD and obesity, dyslipidemia and insulin

resistance, all of which are features of the so-called metabolic syndrome [3,4].

People with the metabolic syndrome are at increased risk of type 2 diabetes (DM2)

and cardiovascular and all-cause mortality [5]. In the majority of cases, NAFLD

causes asymptomatic elevation of liver enzyme levels [6], including alanine

aminotransferase (ALT) and aspartate aminotransferase (AST), which are sensitive

indicators of hepatocellular injury [7]. The former is primarily found in the liver, the

latter is also found in other tissues and is a less specific marker of liver integrity

[7,8]. Recent studies have demonstrated that elevated levels of liver enzymes, and

in particular ALT, are associated with obesity [9], insulin resistance and DM2 [10].

Recent prospective studies found that levels of ALT predicted incident DM2,

independently of the classical risk factors [11] or changes in obesity [10].

Endothelial dysfunctions are an early event in the development of atherosclerosis

[12,13]. One of these functions, i.e. the nitric oxide or endothelium-dependent

regulation of vascular tone can be assessed by measuring vasodilatation of the

brachial artery in response to ischemia-induced increases in blood flow (flow-

mediated dilation; FMD) [14]. The brachial FMD is closely correlated to coronary

endothelial function [15], and endothelial dysfunction predicts future

cardiovascular events [16].

To date, there are no studies in subjects with DM2, assessing the association

between liver enzyme plasma levels and the severity of insulin resistance, and

vascular endothelial dysfunction. The aim of the present study was to assess the

association between alanine aminotransferase (ALT) and aspartate

aminotransferase (AST) levels with whole-body insulin sensitivity and brachial-artery

FMD in normotriglyceridemic subjects with DM2.

CHAPTER 7

116

MATERIALS AND METHODS

Subjects

Patients with DM2 were recruited from shared care projects in Amsterdam and

Hoorn, and via local newspaper advertisements. The following inclusion criteria

were used: age between 40 and 70 years, DM2 as diagnosed by a physician

according to the WHO-criteria [17], HbA1c <8.5% and fasting triglyceride

concentrations as defined as a mean value from three measurements of <2.2

mmol/l, current smoking was allowed. Pregnancy, the premenopausal state

(defined as the presence of menstrual bleeds), the use of lipid lowering drugs or

insulin, alcohol abuse (see below), and liver impairment (serum ALT>70 U/l, serum

AST>50 U/l) or renal impairment (serum creatinine>130 µmol/l or creatinine

clearance<50 ml/min) were exclusion criteria. Written informed consent was

obtained from all subjects and the Ethics Committee of VU University Medical

Center approved the study-protocol. In total, 293 subjects were invited for

screening, of whom 93 fulfilled the inclusion criteria.

Study Design

The study was performed between 1997 and 1999 and the values for the above

mentioned inclusion and exclusion criteria are those that were in accordance with

the prevailing guidelines and generally accepted at that time. Patients were seen on

6 occasions during a period of 14 weeks. The study consisted of one screening visit

(at week –1) to assess eligibility, 3 follow-up visits (at week 0, 6 and 12), during

which dietary intake was assessed and anthropometry and blood collections were

performed (see below). Finally, patients underwent a hyperinsulinemic euglycemic

clamp and a vascular ultrasound examination during 2 separate visits (between

week 12 and 14).

Alcohol intake

Subjects were asked to record their dietary intake from one day in a week

(morning, afternoon and evening), during a total of 5 occasions, with an interval of

2 weeks between the recorded days. From these recordings, the mean alcohol

intake was calculated as an average of the five recorded days. Subjects with a mean

alcohol intake of more than 40 gram per day were excluded from further analyses.


117

Hyperinsulinemic-euglycemic clamps

Insulin sensitivity was measured using a hyperinsulinemic euglycemic clamp

technique as previously described [18]. In short, following an overnight fast,

subjects were placed in the semi-supine position and two polytetrafluoroethylene

cannulas (Venflon, Viggo Helsiborg, Sweden) were inserted into the antecubital or

antebrachial veins of each arm, one for intermittent blood glucose sampling and

the other for the infusion of the glucose and insulin solutions. A volume of 0.5 mL

insulin (100 U mL-1; Velusolin, Novo Nordisk, Bagsvaerd, Denmark) was diluted to

50 mL with 45 mL 0.9% sodium chlorine solution and 4.5 mL human albumin (20%;

CLB, Amsterdam, The Netherlands). Arterialized blood glucose levels were

measured every 5 minutes with an automated glucose oxidase method (Yellow

Springs Instruments, Yellow Springs, OH, USA) after the start of the insulin infusion.

The initial insulin infusion rate (120 mU/m2/min) was decreased to 40 mU/m2/min

once the glucose level was 5.0 mmol/l, and thereafter the glucose 20% infusion

rate was adjusted to maintain this blood glucose level for 2 hours. The M-value,

expressed as glucose uptake in mg/m2/min, was calculated from the average

glucose infusion rate during the last 60 minutes and expressed per unit of plasma

insulin concentration (M/I-value).

Vascular Ultrasound

A non-invasive ultrasound method for measuring FMD of the brachial artery as

described previously was used [19]. In short, vessel diameter of the right brachial

artery was measured with an ultrasound system (Ultramark IV, ATL, Bothell, WA,

USA) in combination with a vessel movement detector system (Wall Track System,

Neurodata, Bilthoven, The Netherlands). Measurements were carried out by a single

experienced ultrasound operator (RJAMvD) in a quiet room, in fasted subjects after

a 15-minute rest in the supine position. An ischemia-induced increase in blood flow

by releasing a blood pressure tourniquet that had been inflated for 4 minutes at

the fore-arm to 100 mmHg above systolic blood pressure, which results in an

increase in shear stress, was used as the stimulus for FMD. Nitroglycerin (0.4 mg

sublingually) was used to obtain nitroglycerin mediated dilation (NMD). FMD was

calculated as the percent change in diameter post-occlusion of the brachial artery

at 60 seconds relative to the measurement at baseline before blood pressure

tourniquet inflation {[(response x baseline)/baseline]e100}. NMD was calculated as

the percent change in diameter before and 5 minutes after administration of the

nitroglycerin.

CHAPTER 7

118

Anthropometric measures and laboratory analysis

A detailed history, in particular regarding alcohol intake and smoking status, was

obtained from each participant. Furthermore height, weight, and waist and hip

circumference were measured with the subjects wearing only light clothes. Waist

circumference was measured at the level midway between the lowest rib margin

and the iliac crest, and hip circumference was measured at the widest level over the

greater trochanters. Body mass index (BMI; weight divided by height squared; kg m-

2) was calculated. Serum cholesterol and HDL-cholesterol were measured with the

CHOD-PAP-method (Roche, Mannheim, Germany). Serum triglycerides were

measured with the GPO-PAP-method (Roche, Mannheim, Germany). LDL-cholesterol

was calculated according to the Friedewald-formula [20]. ALT and AST levels were

determined by enzymatic assays (Roche, Mannheim, Germany). Plasma glucose was

determined with a glucose oxidase method (Granutest, Merck, Darmstadt,

Germany). Glycated hemoglobin (HbA1c) was measured with ion exchange high

performance liquid chromatography (Modular Diabetes Monitoring System, BioRad

Lab, Veenendaal, The Netherlands). Insulin was measured with a double antibody

radio-immunoassay (SP21, Linco Research, St. Louis, USA). Pro-insulin was

measured with an immunometric assay (Dako Diagnostics Ltd., Cambridgeshire,

UK). C-reactive protein (CRP) was measured by an immunoturbidimetric assay (Tina-

quant CRP, Boehringer Mannheim, Mannheim Germany)(reference range: 0-8 mg/l).

Statistical analysis

Analyses were performed by SPSS for Windows 11.0.1 (SPSS Inc. Chicago, IL). Data

are mean (standard deviation) or percentage. Reported values for serum lipids,

glucose, and HbA1c are the mean value of three assessments, which were

performed on the morning of inclusion, on the morning of week 6 and on the

morning of week 12 after an overnight fast; all other laboratory assessments were

determined once. Pearson’s correlation coefficients were calculated to evaluate

relationships between variables. Multiple linear regression analyses were performed

to investigate the associations between ALT as determinant and insulin sensitivity

(M/I-value) and FMD as outcome variables. In the respective regression models for

M/I-value and FMD, adjustments were made for age and sex (Model 1 and Model 4,

respectively). Subsequently, these models were adjusted for AST levels, yielding

Model 2 and Model 5, respectively. Model 3, which had insulin sensitivity as

dependent variable, was Model 2, with additional correction for BMI, since former

studies have shown that obesity was a confounder of the association between ALT


119

and insulin resistance [9,10]. In Model 6, which had FMD as outcome variable,

Model 5 was adjusted for insulin sensitivity since previous investigations found

insulin sensitivity to be an important determinant of FMD [21]. A two-sided P-value

<0.05 was considered to indicate statistical significance.

RESULTS

Patient characteristics

Subjects with missing data on insulin sensitivity and/or patients with missing data

on FMD were excluded (n=26). Three subjects were excluded because of an alcohol

intake of more than 40 grams per day or missing data on alcohol intake.

Table 7.1. Clinical characteristics

N 64

Men / women 35 / 29

Age (years) 61.8 ± 7.2 (44 – 70)

Duration of DM2 (years) 4.7 ± 5.6 (0.3 – 33.6)

Body mass index (kg/m2) 28.5 ± 5.3

Waist (cm) male / female 101 ± 12.9 / 97.7 ± 13.7

Hypertension (%)* 42.2

Systolic blood pressure (mmHg) 144 ± 17

Diastolic blood pressure (mmHg) 83 ± 8

Use of alcohol (%) 65.6

Alcohol intake (g/d) 10.0 ± 11.2

Cigarette smoking (%) 15.5

C-reactive protein (mg/l) 2.6 ± 2.6 (0.12-11.9)

Values are mean ± SD, percentage or (range)

*Defined as RR > 160/90 mmHg and/or use of anti-hypertensive medication.

Finally, 64 subjects were included in the present analyses and their clinical and

biochemical characteristics are listed in Tables 7.1 and 7.2. On average, patients

were normotriglyceridemic and had good glycemic control. Forty-two subjects were

treated by sulfonylurea drugs, 18 by metformin and 2 used α-glucosidase

inhibitors (or combinations thereof).

CHAPTER 7

120

Table 7.2. Metabolic and vascular characteristics

Glucose and insulin metabolism

Fasting glucose concentration (mmol/l) 8.0 ± 1.7

Fasting insulin concentration (pmol/l)* 75.8 ± 43.0

Fasting pro-insulin concentration (pmol/l) 17.8 ± 10.6

M/I-value (mg/m2/min/pmol/l) 0.33 ± 0.17

HbA1c (%) 6.2 ± 0.8

Lipid metabolism

Triglycerides (mmol/l) 1.3 ± 0.4

Total cholesterol (mmol/l) 5.6 ± 0.9

HDL-cholesterol (mmol/l) 1.29 ± 0.3

LDL-cholesterol (mmol/l) 3.7 ± 0.9

Vascular function of the brachial artery

Flow-mediated dilation (%) 6.3 ± 6.0

Nitroglycerin-mediated dilation (%)† 12.4 ± 8.1

Liver enzyme levels‡

Alanine aminotransferase (U/l) 15.0 ± 7.5 (4 - 48)

Aspartate aminotransferase (U/l) 10.6 ± 2.6 (5 - 17)

Values are mean±SD or (range)

*n=62 †n=63 ‡ Reference ranges: ALT<35 U/l; AST<25 U/l

Liver enzyme correlations

Neither ALT nor AST were correlated with the reported amount of alcohol intake

(r=-0.05; P=0.7 and r=-0.11; P=0.4, respectively) or smoking status (r=0.18; P=0.12

and r=0.07; P=0.5, respectively) and additional corrections for alcohol intake or

smoking status did not change the presented correlations. Table 7.3 lists the

correlations between ALT and AST and anthropometric, metabolic and vascular

parameters. ALT, but not AST, was weakly but significantly correlated with waist,

hip, fasting insulin, fasting pro-insulin, M/I-value and NMD. The association

between ALT and FMD tended to be significant (P=0.08). AST was weakly but

significantly correlated with fasting blood glucose levels only.


121

Table 7.3. Correlations among hepatic enzymes and selected anthropometric,

metabolic and vascular parameters

ALT AST

r P r P

Anthropometry

BMI 0.13 0.31 -0.06 0.63

Waist 0.27 0.03 -0.003 0.98

Hip 0.26 0.03 0.08 0.51

Insulin and glucose metabolism

Fasting glucose 0.03 0.83 -0.27 0.03

Fasting insulin* 0.43 <0.001 0.20 0.12

Fasting pro-insulin 0.39 0.006 0.09 0.49

M/I-value -0.29_ 0.02 -0.08 0.53

Vascular function

FMD -0.22 0.08 -0.06 0.63

NMD -0.27 0.04 -0.16 0.22

r, correlations by Pearson; FMD, flow mediated dilation; NMD, nitroglcyerin mediated dilation. *n=62

Associations of liver enzymes with insulin resistance and endothelial function

Multiple linear regression models were used to study the association of ALT with

M/I-values and FMD. In multiple linear regression analyses, a significant association

of insulin sensitivity with ALT after adjustment for AST (Model 2, Table 7.4) and

adjustment for BMI (Model 3, Table 7.4) was found. The age- and gender adjusted

association of FMD and ALT tended to be significant (P=0.06; Model 4, Table 7.4),

was significant after adjustment for AST levels (P=0.045; Model 5, Table 7.4), and

remained significant after adjustment for insulin sensitivity (P=0.03; Model 6, Table

7.4).

CHAPTER 7

122

Table 7.3. Correlations among hepatic enzymes and selected anthropometric,

metabolic and vascular parameters

ALT AST

r P r P

Anthropometry

BMI 0.13 0.31 -0.06 0.63

Waist 0.27 0.03 -0.003 0.98

Hip 0.26 0.03 0.08 0.51

Insulin and glucose metabolism

Fasting glucose 0.03 0.83 -0.27 0.03

Fasting insulin* 0.43 <0.001 0.20 0.12

Fasting pro-insulin 0.39 0.006 0.09 0.49

M/I-value -0.29_ 0.02 -0.08 0.53

Vascular function

FMD -0.22 0.08 -0.06 0.63

NMD -0.27 0.04 -0.16 0.22

r, correlations by Pearson; FMD, flow mediated dilation; NMD, nitroglcyerin mediated dilation. *n=62

Table 7.4. Regression models of the associations of ALT with insulin-sensitivity

(M/I-value) and vascular dysfunction (FMD)

Model* b 95% CI P

M/I-value (*100)

Model 1 -0.73 -1.3 to -0.18 0.010

Model 2 -0.99 -1.7 to -0.30 0.006

Model 3 -0.76 -1.4 to -0.08 0.029

FMD

Model 4 -0.20 -0.40 to +0.08 0.060

Model 5 -0.26 -0.52 to -0.006 0.045

Model 6 -0.31 -0.58 to -0.03 0.030

b, unstandardized regression coefficient; CI, confidence interval; FMD, flow-mediated dilation *Model 1: adjusted for age and gender; Model 2: as Model 1 with additional adjustment for AST; Model 3: as

Model 2 with additional adjustment for BMI; Model 4: adjusted for age and gender; Model 5: as Model 4 with

additional adjustment for AST; Model 6: as Model 5 with additional adjustment for insulin sensitivity (M/I-

value)


123

DISCUSSION We observed that circulating ALT was inversely associated with whole-body insulin

sensitivity, measured with the hyperinsulinemic euglycemic clamp, in

normotriglyceridemic subjects with DM2. Moreover, we demonstrated a negative

association between plasma ALT levels and vascular endothelial function, which

was measured as FMD.

High ALT levels are associated with fatty liver disease and might therefore be

considered as a feature of the metabolic syndrome. Furthermore, elevated ALT, but

not AST levels, were related to hepatic insulin resistance in normoglycemic subjects

at high risk of DM2, and in these subjects, elevated ALT predicted the development

of DM2 [10]. In some studies, the prospective relationship between ALT and DM2

reflected the cross-sectional association with obesity and insulin resistance [9,10],

whereas others found the predictive effect of ALT to be independent of obesity and

body fat distribution [22]. In the present study, the association between ALT and

whole-body insulin sensitivity was independent of BMI.

The pathogenesis of NAFLD is not yet fully understood but oxidative stress and

lipid peroxidation resulting from excessive lipid accumulation (lipotoxicity) are

regarded as important pathophysiological factors [23-25]. Central obesity and the

inappropriate suppression of lipolysis arising from insulin resistance, result in an

increased flux of non-esterified or free fatty acids (FFA) and an increased output of

triglyceride-rich VLDL particles from the liver. Due to the increased triglyceride/FFA

load, in the presence of impaired glucose utilization, non-adipose tissues, including

liver, muscle and the pancreatic β cells, accumulate triglycerides [26]. In these

tissues, intracellular triglyceride overload increases the pool of long-chain acyl Co-

A, thus providing abundant substrate for nonoxidative metabolic pathways and the

formation of toxic metabolites and also causing mitochondrial dysfunction and

subsequent production of reactive oxygen species. These lipid-induced changes

ultimately lead to organ dysfunction and cell apoptosis.

At present there are no studies addressing the value of ALT as a risk marker once

DM2 has developed. In addition, since hypertriglyceridemia is one of the hall-marks

of the metabolic syndrome [27], demonstration of high plasma ALT levels in

normotriglyceridemic DM2 subjects may be of additional value in predicting high

risk of microvascular and/or macrovascular complications. Abnormal liver function

tests are more common in DM than in non-DM populations as well as in subjects

CHAPTER 7

124

with DM2 than in those with DM1 [28]. In DM1, elevated liver enzymes were

associated with diabetic complications such as retinopathy and neuropathy,

independently of alcohol consumption, BMI and glycemic control [29]. To our

knowledge, at present no such data have been reported for DM2 subjects. Here, we

show an association between ALT levels and impaired endothelial function in

normotriglyceridemic subjects with DM2. Previously, the presence of generalized

endothelial dysfunction in insulin-resistant subjects and patients with DM2 was

proposed to explain the excess cardiovascular morbidity and mortality in these

populations [21]. The mechanisms underlying this endothelial dysfunction include

insulin resistance at the level of the vascular wall, in part due to elevated circulating

FFA levels, causing a reduction in insulin-mediated nitric oxide production, as well

as insulin- resistance associated induction of inflammatory pathways and oxidative

stress [30,31]. Collectively, these mechanisms decrease the bioavailability of NO

leading to loss of the anti-atherogenic properties of the vascular endothelium

including impairment of NO-mediated vasodilatation. Indeed, the significant

association between ALT and FMD did not change after adjustment for insulin

resistance, indicating that the association of ALT with FMD is not only mediated by

insulin resistance but that other factors, including oxidative stress and

inflammation, may also play a role. This latter suggestion is in accordance with a

recent study that demonstrated that ALT is associated with CRP-levels, indicating

that a low-grade inflammatory process is involved in the pathogenesis of NAFLD

[32].

Our study has some limitations. First, the study population consisted of patients

with relatively well-controlled DM2 with regard to glycaemia and triglyceridemia

and therefore, the observed associations may be an underestimation of the

relationships in diabetic subjects with poor metabolic control. Second, we used

elevation of liver enzyme levels as an indicator of NAFLD instead of biopsies, which

are regarded as the gold standard. Third, although we addressed alcohol as major

potential confounder in this study, we could not account for other factors that may

cause liver enzyme elevations, including iron-status [33], infections and/or

hepatotoxic medication [34]. However, we acknowledge that the cut-off value of 40

grams of alcohol per day is slightly higher that the most recently suggested cut-off

value of 20 grams per day for women and 40 grams per day for men [35]. Fourth,

the study had a cross-sectional design and the observed associations cannot be

interpreted as a reflection of causal relationships. Finally, although we assessed

whole-body insulin resistance and not hepatic insulin resistance, the positive


125

association of ALT with fasting insulin values might indicate that clearance of

insulin in the liver is less effective [36,37].

In summary, in relatively well-controlled patients with DM2, plasma ALT levels are

related to a decline in insulin-sensitivity and an impairment of conduit vessel

endothelial function. The findings might suggest that elevated ALT is a marker of

liver fat also in normotriglyceridemic DM2. Additional prospective studies are

needed to confirm these results in other study populations and further clarify the

underlying mechanisms.

CHAPTER 7

126

REFERENCES

1. Angulo P. Nonalcoholic Fatty liver disease. N Engl J Med 2002;346:1221-31.

2. Matteoni CA, Younossi ZM, Gramlich T, et al. Nonalcoholic fatty liver disease: a

spectrum of clinical and pathological severity. Gastroenterology 1999;116:1413-19.

3. Powell EE, Cooksley WG, Hanson R, et al. The natural history of nonalcoholic

steatohepatitis: a follow-up study of forty-two patients for up to 21 years. Hepatology

1990;11:74-80.

4. Marchesini G, Bugianesi E, Forlani G, et al. Nonalcoholic fatty liver, steatohepatitis,

and the metabolic syndrome. Hepatology 2003;37:917-23.

5. Isomaa B, Almgren P, Tuomi T, et al. Cardiovascular morbidity and mortality

associated with the metabolic syndrome. Diabetes Care 2001;24:683-89.

6. Daniel S, Ben Menachem T, Vasudevan G, Ma CK, Blumenkehl M. Prospective

evaluation of unexplained chronic liver transaminase abnormalities in asymptomatic

and symptomatic patients. Am J Gastroenterol 1999;94:3010-4.

7. Ellis G, Goldberg DM, Spooner RJ, Ward AM. Serum enzyme tests in diseases of the

liver and biliary tree. Am J Clin Pathol 1978;70:248-58.

8. Reichling JJ, Kaplan MM. Clinical use of serum enzymes in liver disease. Dig Dis Sci

1988;33:1601-14.

9. Lee DH, Ha MH, Christiani DC. Body weight, alcohol consumption and liver enzyme

activity--a 4-year follow-up study. Int J Epidemiol 2001;30:766-70.







Diabetes 2004;53:2855-60.

12. Toborek M, Kaiser S. Endothelial cell functions. Relationship to atherogenesis. Basic

Res Cardiol 1999;94:295-314.

13. Ross R. Atherosclerosis--an inflammatory disease. N Engl J Med 1999;340:115-26.

14. Corretti MC, Anderson TJ, Benjamin EJ, et al. Guidelines for the ultrasound

assessment of endothelial-dependent flow-mediated vasodilation of the brachial

artery: a report of the International Brachial Artery Reactivity Task Force. J Am Coll

Cardiol 2002;39:257-65.

15. Anderson TJ, Uehata A, Gerhard MD, et al. Close relation of endothelial function in

the human coronary and peripheral circulations. J Am Coll Cardiol 1995;26:1235-41.

16. Suwaidi JA, Hamasaki S, Higano ST, et a.. Long-term follow-up of patients with mild

coronary artery disease and endothelial dysfunction. Circulation 2000;101:948-54.

17. World Health Organization. WHO Expert Committee on Diabetes Mellitus. Report of a

WHO Study Group. Technical Report Series 27. 1985.


127

18. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying

insulin secretion and resistance. Am J Physiol Endocrinol Metab 1979;237:214-23.

19. Lambert J, Aarsen M, Donker AJM, Stehouwer CDA. Endothelium-dependent and -

independent vasodilation of large arteries in normoalbuminuric insulin-dependent

diabetes mellitus. Arterioscler Thromb Vasc Biol 1996;16:705-11.




21. Steinberg HO, Baron AD. Vascular function, insulin resistance and fatty acids.

Diabetologia 2002;45:623-34.

22. Ohlson LO, Larsson B, Bjorntorp P, et al. Risk factors for type 2 (non-insulin-

dependent) diabetes mellitus. Thirteen and one-half years of follow-up of the

participants in a study of Swedish men born in 1913. Diabetologia 1988;31:798-805.

23. Day CP, James OF. Steatohepatitis: a tale of two "hits"? Gastroenterology

1998;114:842-5.

24. Wang JT, Liu YL. Non-alcoholic fatty liver disease: the problems we are facing.

Hepatobiliary Pancreat Dis Int 2003;2:334-7.

25. Day CP. Pathogenesis of steatohepatitis. Best Pract Res Clin Gastroenterol

2002;16:663-78.

26. Unger RH. Lipotoxic diseases. Annu Rev Med 2002;53:319-36.

27. Grundy SM, Brewer HB, Jr., Cleeman JI, Smith SC, Jr., Lenfant C. Definition of

metabolic syndrome: Report of the National Heart, Lung, and Blood

Institute/American Heart Association conference on scientific issues related to

definition. Circulation 2004;109:433-38.



29. Arkkila PE, Koskinen PJ, Kantola IM, et al. Diabetic complications are associated with

liver enzyme activities in people with type 1 diabetes. Diabetes Res Clin Pract

2001;52:113-8.

30. Dandona P, Aljada A, Chaudhuri A, Bandyopadhyay A. The potential influence of

inflammation and insulin resistance on the pathogenesis and treatment of

atherosclerosis-related complications in type 2 diabetes. J Clin Endocrinol Metab

2003;88:2422-9.

31. Evans JL, Goldfine ID, Maddux BA, Grodsky GM. Oxidative stress and stress-activated

signaling pathways: a unifying hypothesis of type 2 diabetes. Endocr Rev

2002;23:599-622.




CHAPTER 7

128

33. Hsiao TJ, Chen JC, Wang JD. Insulin resistance and ferritin as major determinants of

nonalcoholic fatty liver disease in apparently healthy obese patients. Int J Obes Relat

Metab Disord 2004;28:167-72.

34. Kaplowitz N. Drug-induced liver injury. Clin Infect Dis 2004;38 (Suppl. 2):S44-8

35. Cortez-Pinto H. Non-alcoholic fatty liver disease/non-alcoholic steatohepatitis

(NAFLD/NASH): diagnosis and clinical course. Best Pract Res Clin Gastroenterol

2004;18:1089-104.

36. Valera Mora ME, Scarfone A, Calvani M, Greco AV, Mingrone G. Insulin clearance in

obesity. J Am Coll Nutr 2003;22:487-93.

37. Strang BD, Bertics SJ, Grummer RR, Armentano LE. Relationship of triglyceride

accumulation to insulin clearance and hormonal responsiveness in bovine

hepatocytes. J Dairy Sci 1998;81:740-7.

129

Chapter 8

ALANINE AMINOTRANSFERASE PREDICTS CORONARY HEART DISEASE EVENTS: A TEN YEAR FOLLOW-UP OF THE HOORN STUDY



Giel Nijpels2,3

Lex M. Bouter2

Coen D. A. Stehouwer4

Robert J. Heine1,2

Michaela Diamant1

From the Departments of 1Endocrinology/Diabetes Center, 3General Practice, the

2EMGO Institute, VU University Medical Center, Amsterdam, The Netherlands, and

the 4Department of Internal Medicine, Academic Hospital Maastricht, Maastricht,

The Netherlands.

Atherosclerosis 2007;191:391-6

Copyright © 2006 Elsevier Ireland Ltd.

Reprinted with permission from Elsevier.

CHAPTER 8

130

ABSTRACT

Alanine aminotransferase (ALT) is a marker of nonalcoholic fatty liver disease

(NAFLD) and predicts incident type 2 diabetes mellitus (DM2). Recently, ALT was

shown to be also associated with endothelial dysfunction and carotid

atherosclerosis. We studied the predictive value of ALT for all cause mortality,

incident cardiovascular disease (CVD) and coronary heart disease (CHD) events in a

population-based cohort of Caucasian men and women aged 50 to 75 years, at

baseline. The ten-year risk of all-cause mortality, fatal and non-fatal CVD and CHD

events in relation to ALT was assessed in 1439 subjects participating in the Hoorn

study, using Cox survival analysis. Subjects with prevalent CVD/CHD and missing

data were excluded. As compared with the first tertile, the age- and sex-adjusted

hazard ratios (95 percent confidence intervals) for all-cause mortality, CVD events

and CHD events were 1.30 (0.92-1.83), 1.40 (1.09-1.81) and 2.04 (1.35-3.10),

respectively, for subjects in the upper tertile of ALT. After adjustment for

components of the metabolic syndrome and traditional risk factors, the association

of ALT and CHD events remained significant for subjects in the third relative to

those in the first tertile, with a hazard ratio of 1.88 (1.21-2.92) and 1.75 (1.12-

2.73), respectively. In conclusion, the predictive value of ALT for coronary events,

seems independent of traditional risk factors and the features of the metabolic

syndrome in a population based cohort. Further studies should confirm these

findings and elucidate the pathophysiological mechanisms.

ALANINE AMINOTRANSFERASE AND CORONARY HEART DISEASE EVENTS

131

INTRODUCTION

Recently, the role of non-alcoholic fatty liver disease (NAFLD) in the pathogenesis of

type 2 diabetes mellitus (DM2) has gained much interest. Several studies have

demonstrated that NAFLD is associated with the components of the metabolic

syndrome (MetS) and DM2 [1,2], and fatty liver is considered to be the hepatic

component of the MetS [3,4]. Circulating concentrations of the liver transaminases,

alanine aminotransferase (ALT) and aspartate aminotransferase (AST) have been

used as markers of NAFLD. Of these two liver enzymes, ALT appears to be the best

marker of liver fat accumulation and is positively correlated with liver fat measured

by magnetic resonance (MR) proton spectroscopy [5]. In cross-sectional and

prospective studies ALT is associated with DM2 [6,7]. Collectively, these data

suggest that hepatic steatosis might play a role in the pathogenesis of DM2, most

probably by contributing to the development of hepatic insulin resistance resulting,

among others, in increased hepatic gluconeogenesis and overproduction of

triglyceride-rich lipoproteins. Subjects with features of the MetS are not only at

increased risk of DM2, but also at risk of developing cardiovascular (CVD) and

coronary heart disease (CHD) [8].

In view of these observations, and the recently described associations between

both NALFD or its marker ALT and vascular structural and functional properties

[9,10], we studied the predictive value of ALT for all-cause mortality, CVD and CHD

events in a population-based study of diabetes and related complications in

Caucasian men and women aged 50 to 75 years.

METHODS

Study population

Subjects were participants in the Hoorn Study, a prospective population-based

cohort study of glucose metabolism and diabetes complications. The study

population and research design have been described in detail previously [11]. In

short, in 1989, a random sample of all men and women aged 50 to 75 was taken

from the municipal registry of the medium-sized town of Hoorn in The Netherlands.

Of the 3552 individuals, who where invited to take part in the study, 2540 agreed

to participate (71.5%). Baseline data, including a 75-g oral glucose tolerance test

(OGTT), were collected from October 1989 through February 1992. After excluding

CHAPTER 8

132

56 non-Caucasians the cohort consisted of 2484 men and women. For the present

study we excluded 612 subjects with missing information on morbidity during

follow-up, because they did not provide written permission to accesses their

medical records or they had moved out of Hoorn. Subjects with prevalent CVD/CHD

and missing data on variables at baseline were excluded for the analysis, resulting

in 1439 subjects. The Ethical Review Committee of VU University Medical Centre

approved the Hoorn Study and written informed consent was obtained from all

participants.

Baseline Examination

Serum ALT enzyme activity was measured according to the method of the

International Federation of Clinical Chemistry from 1985, and expressed as U/L

[12]. Since the reference range and cut-off values for ALT are controversial [13,14],

we divided ALT into tertiles, instead of using a cut-off value to define abnormality.

The OGTT was performed between 08.00 and 10.00h and subjects were asked not

to drink alcohol from 17.00h and to fast (except for drinking water) from 22.00h

the evening before the measurements. Blood samples were collected before and 2h

after ingestion of the glucose load. Plasma glucose was measured with the glucose

dehydrogenase method (Merck, Darmstadt, Germany). The glucose tolerance

categories were defined according to the 1999 criteria of the World Health

Organization. Immuno-specific insulin was measured in serum with a double-

antibody RIA (antibody SP21, Linco Research, St. Louis, USA). Insulin resistance was

estimated by the homeostasis model assessment for insulin resistance (HOMA-IR),

calculated as fasting insulin (in IU/l ) x fasting glucose (mmol/l)/22.5 [15]. Glycated

haemoglobin (HbA1c) was measured with ion exchange high performance liquid

chromatography (Modular Diabetes Monitoring System, BioRad Lab, Veenendaal,

The Netherlands; reference range 4.3-6.1%). Fasting triglycerides, total and HDL-

cholesterol were determined in serum by enzymatic techniques (Boehringer

Mannheim, Germany). LDL-cholesterol was calculated according to the Friedewald-

formula, except in subjects with triglycerides>4.5 mmol/l. Seated systolic and

diastolic blood pressure was measured at the right arm after a 5-minute rest with a

random-zero mercury sphygmomanometer (Hawksley-Gelman, Lancing, UK). The

average of two measurements was used for the analyses. Weight and height were

measured in subjects wearing light clothes only, and the body mass index (BMI)

was calculated as weight divided by height squared (kg/m2). Waist circumference

was measured at the midpoint between the lowest rib and the iliac crest, hip

circumference was measured at the widest level over the greater trochanters, and


133

the mean value of two measurements was used in the analysis. Information on

alcohol intake was assessed by a validated semi-quantitative food frequency

questionnaire [16]. Subjects were asked about their drinking habits and the weekly

number of glasses of alcoholic drinks consumed. This information was converted to

alcohol intake (g/day) using a computerized version of the Dutch Food

Composition Table. Physical activity was assessed by questionnaire as described

before [11]. Information on history of cardiovascular disease was assessed by the

Rose questionnaire. Smoking status was assessed by questionnaire and categorized

as never, former and current smoker.

Follow-up

The registration of follow-up data of the Hoorn Study participants is regularly

updated using the municipal register of the city of Hoorn providing information

about the vital status of the participants. Information of causes of death and non-

fatal events were obtained from medical records of general practitioners and from

the local hospital. Causes of death were coded according to the ICD-9. CVD was

defined as documented angina pectoris (chest pain followed by coronary artery

bypass surgery or angioplasty, or in the presence of >50 % stenosis or ECG

changes, i.e. ST-segment elevation ≥ 1 mm or ST-segment depression ≥0.5 mm, or

positive exercise test), myocardial infarction (in the presence of at least 2 of the

following: typical pain, elevated enzymes and/or ECG changes), congestive heart

failure (in the presence of at least 2 of the following: shortness of breath,

cardiomegaly, dilated neck veins or 1 of the former in the presence of edema or

tachycardia), stroke or transient ischaemic attack (sudden onset of symptoms,

neurological symptoms or change of consciousness), peripheral disease (both

symptomatic and a-symptomatic: by procedure or typical pain accompanied by

stenosis or ankle arm blood pressure ratio <0.9 or positive vascular stress test). In

fatal cases, CVD was defined with ICD codes 390-459 (Diseases of the circulatory

system) or 798 (sudden death, cause unknown), because sudden death in general

is of CVD origin [17]. CHD was defined as documented myocardial infarction and

angina pectoris, with the aforementioned criteria. Fatal CHD cases were defined

with ICD coded 410-414 (ischaemic heart disease). Follow-up time was calculated

as the time between the date of the baseline examination and the date of the first

event or January 1, 2000.

CHAPTER 8

134


Analyses were performed with SPSS 11.0.5 software. Subjects were categorized

according to ALT tertiles. Differences between tertiles were tested with analysis of

variance (ANOVA) for continuous variables and by the χ-square test for proportions.

Hazard ratios and 95% confidence intervals of baseline ALT for all-cause mortality,

incident CVD and CHD events were estimated from multivariable Cox proportional

hazard analysis. The first models were adjusted for age and sex only. In the

subsequent models we added the possible mediating or confounding factors one

by one. In the final models, beside adjustments for age, sex, alcohol-intake,

smoking and physical activity, we corrected for the traditional CVD risk factors, i.e.

glucose tolerance status, systolic blood pressure, HbA1c, LDL-cholesterol, BMI

(Table 2, Model 2) as well as for the components of the metabolic syndrome, as

defined by the Adult Treatment Panel III (NCEP) (Table 2, Model 3), respectively.

Two-sided P values < 0.05 were considered to indicate statistical significance.

RESULTS

Baseline characteristics of the participants

The mean age of the 1439 participants (651 men and 788 women) was 60.9 (7.2)

years. Table 8.1 shows the baseline characteristics of the participants divided into

ALT tertiles. Subjects in the highest ALT tertile had a higher BMI, as well as waist

and hip circumference. Also, they had higher fasting and 2 hour post-load plasma

glucose levels, fasting insulin levels, total cholesterol concentrations, and higher

blood pressure. LDL-cholesterol tended to be different across the tertiles, whereas

HbA1c was similar in all tertiles. The highest ALT tertile consisted of a higher

number of subjects with impaired glucose metabolism and DM2.

Association of ALT and all-cause mortality, CVD and CHD events

During the 10-year follow-up, 174 of the total of 1439 subjects died, 355 CVD

events occurred of which 75 where fatal and 129 CHD events occurred of which 16

were fatal. Table 8.2 shows the hazard ratios of the third relative to the first ALT

tertile for all-cause mortality, CVD and CHD events. The age and sex adjusted risk

for all-cause mortality, CVD events and CHD events were 1.30 (0.92-1.83), 1.40

(1.09-1.81) and 2.04 (1.35-3.10), respectively, for the subjects in the upper ALT

tertile compared to those in the first tertile. The association between ALT and all-


135

cause mortality was not statistically significant and the association could be

explained by the traditional CVD risk factors and/or the components of the

metabolic syndrome (Table 8.2, Model 2 and Model 3). The significant association

of ALT with CVD events disappeared after adjustment for the classical CVD risk

factors and/or the components of the metabolic syndrome (Table 8.2, Model 2 and

3). Only the association of ALT and CHD events remained statistically significant

after adjustment for CVD risk factors as well as for the features of the metabolic

syndrome. The association of ALT and CHD events was explained by BMI, waist and

to a lesser extent by total cholesterol and triglycerides. Fasting insulin and HOMA-

IR did not substantially alter any of the associations presented (Table 8.2).

Additional adjustment for baseline HDL in model 2 and additional adjustment for

baseline LDL in model 3 (both HDL and LDL in one model) did not change the

presented associations of ALT with CHD (data not shown).

ALT and CHD events: the effect of gender, alcohol intake and diabetes

In order to establish the robustness of the observed associations, we performed

additional analyses, focusing on the role of gender, alcohol intake and DM2.

Stratified analysis for men and women separately yielded similar results (data not

shown). Since DM2 carries a greater risk of CHD and CVD, and its association with

elevated ALT, we subsequently excluded the 119 DM2 subjects from the analysis,

which resulted in an attenuation of the association of ALT with CHD events. After

adjustment for traditional CVD risk factors, the association became borderline

significant (HR 1.65 [0.97-2.79]), whereas it remained significant after adjustment

for the components of the metabolic syndrome (HR 1.78 [1.05-3.04]).

In order to detail the effect of alcohol on the association of ALT with CHD-events,

we excluded 232 subjects (77 women and 155 men) who had an average daily

alcohol intake of more than 20 g. Exclusion of these subjects attenuated the

association, however, ALT remained a predictor for CHD after adjustment for

classical CVD risk factors (HR 1.72 [1.02-2.90]) and for the components of the

metabolic syndrome (HR 1.67 [0.99-2.80]).

CHAPTER 8

136

Table 8.1. Baseline characteristics of the participants (N=1439) stratified by ALT

tertiles*

Variable Alanine aminotransferase activity (U/L)

Mean (SD), median (interquartile

range), or percentage

1

N=551

2

N=420

3

N=468

P

ALT (U/L)† 12 (1-14) 17 (15-20) 26 (21-143)

Age (years) 62.9 (7.4) 60.7 (7.1) 59.7 (6.8) <0.001

Male (%) 30.9 48.6 59.2 <0.001

Body mass index (kg/m2) 25.7 (3.4) 26.1 (3.2) 27.4 (3.3) <0.001

Waist (cm) 87.0 (10.1) 89.6 (10.3) 94.2 (10.1) <0.001

Hip (cm) 101.3 (6.7) 100.9 (6.2) 102.9 (6.4) <0.001

Fasting glucose (mmol/l) 5.5 (1.1) 5.7 (1.4) 5.9 (1.5) <0.001

2h post-load glucose (mmol/l) 5.6 (2.3) 5.8 (3.0) 6.5 (3.4) <0.001

Fasting insulin (pmol/l) 79.8 (48.2) 82.6 (49.9) 96.1 (58.0) <0.001

HbA1c (%) 5.4 (0.7) 5.5 (0.8) 5.5 (0.9) 0.15

Impaired glucose metabolism (%) 12.0 15.0 21.6 <0.001

Type 2 diabetes mellitus (%) 6.0 7.6 8.2 0.03

Plasma lipids (mmol/l)

Total-cholesterol 6.6 (1.1) 6.5 (1.1) 6.7 (1.2) <0.001

LDL-cholesterol 4.6 (1.0) 4.5 (1.0) 4.7 (1.1) 0.051

HDL-cholesterol 1.38 (0.36) 1.38 (0.37) 1.30 (0.36) <0.001

Triglycerides 1.4 (0.9) 1.4 (0.7) 1.8 (1.1) <0.001

Systolic BP (mmHg) 133 (20) 134 (20) 137 (20) <0.001

Diastolic BP (mmHg) 80 (11) 82 (10) 84 (10) <0.01

Smoking, current (%) 35.2 27.9 26.3 <0.001

Alcohol intake (g/day) 2.2 (0-10) 5.0 (0-15) 10 (2-17) <0.01

*Abbreviations: ALT, alanine aminotransferase; SD, standard deviation; BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein

†median (minimum-maximum value)


137

Table 8.2. Hazard ratios of the second and third compared to the first tertile of ALT

in relation to all-cause mortality, CVD and CHD events*

Model All-cause

mortality

CVD-events CHD-events

HR (95%CI) HR (95%CI) HR (95%CI)

Events Tertile 174 355 129

Model 1† 2nd 0.74 (0.49-1.10) 1.02 (0.78-1.33) 0.94 (0.57-1.53)

3rd 1.30 (0.92-1.83) 1.40 (1.09-1.81) 2.04 (1.35-3.10)

Model 1 + alcohol-intake 2nd 0.74 (0.50-1.10) 1.02 (0.78-1.33) 0.93 (0.57-1.52)

3rd 1.30 (0.92-1.84) 1.39 (1.08-1.80) 2.03 (1.38-3.09)

Model 1 + BMI 2nd 0.72 (0.48-1.08) 0.98 (0.75-1.28) 0.90 (0.55-1.47)

3rd 1.23 (0.85-1.76) 1.27 (0.98-1.65) 1.82 (1.18-2.81)

Model 1 + waist 2nd 0.70 (0.47-1.05) 0.96 (0.73-1.26) 0.89 (0.55-1.46)

3rd 1.15 (0.80-1.65) 1.24 (0.96-1.62) 1.80 (1.17-2.77)

Model 1 + triglycerides 2nd 0.73 (0.49-1.09) 1.02 (0.78-1.33) 0.97 (0.57-1.53)

3rd 1.24 (0.87-1.75) 1.34 (1.04-1.74) 1.97 (1.25-2.99)

Model 1 + total cholesterol 2nd 0.74 (0.49-1.09) 1.02 (0.78-1.32) 0.92 (0.57-1.51)

3rd 1.27 (0.89-1.80) 1.36 (1.05-1.75) 1.89 (1.25-2.88)

Model 1 + fasting glucose 2nd 0.67 (0.45-1.01) 0.96 (0.74-1.26) 0.92 (0.56-1.51)

3rd 1.21 (0.79-1.60) 1.25 (0.97-1.62) 1.98 (1.30-3.02)

Model 1 + fasting insulin 2nd 0.72 (0.48-1.07) 1.03 (0.79-1.34) 0.95 (0.58-1.56)

3rd 1.29 (0.91-1.83) 1.41 (1.09-1.82) 2.06 (1.35-3.14)

Model 1 + HOMA-IR 2nd 0.71 (0.47-1.05) 1.01 (0.78-1.32) 0.95 (0.58-1.56)

3rd 1.20 (0.85-1.72) 1.34 (1.03-1.74) 2.02 (1.32-3.10)

Model 2‡ 2nd 0.75 (0.50-1.12) 0.99 (0.76-1.30) 0.89 (0.54-1.47)

3rd 1.21 (0.83-1.76) 1.25 (0.96-1.64) 1.75 (1.12-2.73)

Model 3§ 2nd 0.71 (0.48-1.06) 0.99 (0.76-1.20) 0.93 (0.57-1.52)

3rd 1.10 (0.77-1.61) 1.22 (0.94-1.60) 1.88 (1.21-2.92)

*Abbreviations: ALT, alanine aminotransferase; CVD, cardiovascular disease; CHD, coronary heart disease; HR,

hazard ratio; CI, confidence interval; BMI, body mass index; HOMA-IR, homeostasis model assessment for

insulin resistance

†Model 1: adjusted for age and sex

‡Model 2: adjusted for age, sex, alcohol-intake, smoking, physical activity, glucose tolerance status, systolic

blood pressure, HbA1c, LDL-cholesterol, BMI

§Model 3: adjusted for age, sex, alcohol-intake, smoking, physical activity, waist, triglycerides, systolic blood

pressure, fasting glucose, HDL-cholesterol

CHAPTER 8

138

DISCUSSION

The main novel finding of this population-based study with well-documented

characterization and follow-up of the participants, was a significant prospective

association of ALT with CHD events, independent of the traditional CVD risk

factors, such as systolic blood pressure, HbA1c, LDL-cholesterol and BMI, and

independent of the components of the NCEP-defined metabolic syndrome. In

addition we confirmed the cross-sectional association of ALT, a marker of NAFLD,

with components of the MetS, including measures of obesity, fasting glucose and

lipids. No significant associations were found between ALT and CVD events or all-

cause mortality, the latter observation is in accordance with a previous study that

found no significant association of ALT with all-cause mortality [18]. The reason

that we did not demonstrate a significant association of ALT and CVD events may

be explained by that fact that the assessment of non-fatal incident CVD may be

more subject to non-differential misclassification than non-fatal incident CHD

events, thus leading to an underestimation of the true associations. Our study is, to

our best knowledge, the first to demonstrate a prospective relation of ALT with

CHD events and extends the results from cross-sectional studies demonstrating

correlations of ALT with CHD risk factors, including the components of the MetS

[9,19]. Bruckert et al found that elevated ALT was associated with CHD risk factors,

including elevated blood pressure, total cholesterol and triglyceride concentrations

in 8,501 hyperlipidemic subjects [19]. We recently reported a positive association

between ALT and endothelial dysfunction measured as brachial flow mediated

vasodilatation. This association was independent of whole-body insulin sensitivity,

indicating that other mechanisms besides insulin resistance, such as inflammation

and/or oxidative stress might contribute to this association [9]. This is in

accordance with the data from our present study, demonstrating that adjustment

with HOMA-IR did not change the age and sex adjusted models to a significant

extent. It should be noted that, although HOMA-IR is a validated surrogate measure

for whole-body insulin sensitivity, it may not be a specific marker of hepatic insulin

resistance. Targher et al showed that healthy male volunteers with ultrasound

documented NAFLD had an increased cIMT relative to those without NAFLD [20].

Since this relation was largely explained by the amount of visceral abdominal fat,

adipose-tissue derived adipocytokines were implicated as a link between fatty liver

and atherosclerosis.


139

Taken together, three mechanisms, which all seem interrelated, may explain the

association between ALT and the risk of coronary atherosclerosis. First, NALFD,

often represented by its marker ALT, are linked with both hepatic and systemic

insulin resistance and various components of the MetS [7,9]. As shown in recent

reports form different populations, adults with the MetS are at increased risk of

CHD, CVD and all-cause mortality [8,21]. Interestingly, however, adjustment for the

components of the MetS did not blunt the association of ALT and CHD, indicating

other contributing mechanisms. In addition, NAFLD is often referred to as the

hepatic component of the MetS, and adjustment for the (other) components of the

MetS may lead to an over adjustment in the association of ALT and CHD, since it is

not yet fully clear whether the other components are cause, consequence or

intermediate factors in the pathophysiology of NAFLD and CHD. Second, the low-

grade systemic and hepatic inflammatory state may link elevated ALT to a high risk

of CHD. In the normal liver, the production of cytokines is absent or minimal

(insignificant), however, various stimuli such as reactive oxygen species, may

induce the production of cytokines, such as tumour-necrosis factor-alpha and

interleukin-6 [22], leading to inflammation and fibrosis. In addition to hepatic low-

grade inflammation, adipose tissue compartments, both in healthy and in obese

subjects are a major source of these cytokines. These cytokines may further

amplify the inflammatory cascade by stimulating hepatic CRP production. A recent

study demonstrated higher CRP levels in subjects with elevated ALT compared to

subjects with normal ALT, even after adjustment for major confounding factors

[23]. Indeed, CRP may be important in linking NAFLD to atherosclerosis since an

elevated plasma CRP level is by now established as an independent risk factor of

CHD and CVD events [24,25]. Third, elevation of ALT in response to fatty

infiltration of the liver may similarly reflect excessive fat deposition in other organs

such as the myocardium and the skeletal muscle. In these non-adipose tissues,

intracellular triglyceride overload increases the pool of long-chain acyl Co-A, thus

providing abundant substrate for non-oxidative metabolic pathways and the

formation of toxic metabolites and also causing mitochondrial dysfunction and

subsequent production of reactive oxygen species and ultimately leading to organ

dysfunction and cell apoptosis [26].

A number of potential limitations of this study should be considered. The

population consisted of Caucasian individuals aged 50 to 75, and caution should

be exercised to generalize our findings to other populations. We excluded subjects

with missing information on nonfatal disease because they did not give permission

to access their hospital records. To study the possibility of selection bias, we

CHAPTER 8

140

repeated the analyses for all-cause mortality using the data from the entire original

Hoorn Study cohort, without prevalent CVD/CHD (n=2007), and the estimates were

almost identical (data not shown). We used, similar to others, ALT levels as a

marker of NAFLD instead of imaging or biopsies, the latter being regarded as the

gold standard. Since serum ALT shows a strong correlation with biopsy proven fatty

liver disease in patients and with liver triglyceride content, measured by MRI/MRS

[27,28], its use as a marker of NAFLD in epidemiological studies seems

appropriate. Although we addressed alcohol as an important confounder in our

study, we could not account for other factors that might have caused a rise in ALT,

such as viral hepatitis infections or hepatotoxic drugs. We did not assess the

association of other liver enzymes such as AST and gamma-glutamyl transferase

with outcome, as data on these markers were not available in the Hoorn Study

cohort. Finally, we did not measure CRP at baseline in the Hoorn Study, since low

grade inflammation may be a contributing mechanism in the association of ALT

and CHD, future studies should address this issue.

The results of the present study show that ALT predicts CHD events, independent

of traditional CVD risk factors and the components of the metabolic syndrome.

Further studies are warranted to confirm these findings and to elucidate the

pathophysiological mechanisms.


141

REFERENCES 1. Gupte P, Amarapurkar D, Agal S, et al. Non-alcoholic steatohepatitis in type 2

diabetes mellitus. J Gastroenterol Hepatol 2004;9:854-8.

2. Younossi ZM, Gramlich T, Matteoni CA, et al. Nonalcoholic fatty liver disease in

patients with type 2 diabetes. Clin Gastroenterol Hepatol 2004;2:262-5.

3. Cortez-Pinto H, Camilo ME, Baptista A, et al. Non-alcoholic fatty liver: another feature

of the metabolic syndrome? Clin Nutr 1999;18:353-8.

4. Knobler H, Schattner A, Zhornicki T, et al. Fatty liver--an additional and treatable

feature of the insulin resistance syndrome. QJM 1999;92:73-9.




2004;47:1360-9.




Diabetes 2004;53:2855-60.




8. Dekker JM, Girman C, Rhodes T, et al. Metabolic syndrome and 10-year cardiovascular

disease risk in the Hoorn Study. Circulation 2005;112:666-73.




10. Brea A, Mosquera D, Martin E, et al. Nonalcoholic fatty liver disease is associated with

carotid atherosclerosis: a case-control study. Arterioscler Thromb Vasc Biol

2005;25:1045-50.



1995;18:1270-3.







2002;137:49-51.



CHAPTER 8

142




16. Grootenhuis PA, Westenbrink S, Sie CM, et al. A semiquantitative food frequency

questionnaire for use in epidemiologic research among the elderly: validation by

comparison with dietary history. J Clin Epidemiol 1995;48:859-68.

17. Kannel WB, Plehn JF, Cupples LA. Cardiac failure and sudden death in the Framingham

Study. Am Heart J 1988;115:869-75.

18. Arndt V, Brenner H, Rothenbacher D, et al. Elevated liver enzyme activity in

construction workers: prevalence and impact on early retirement and all-cause

mortality. Int Arch Occup Environ Health 1998;71:405-12.

19. Bruckert E, Giral P, Ratziu V, et al. A constellation of cardiovascular risk factors is

associated with hepatic enzyme elevation in hyperlipidemic patients. Metabolism

2002;51:1071-6.

20. Targher G, Bertolini L, Padovani R, et al. Relation of nonalcoholic hepatic steatosis to

early carotid atherosclerosis in healthy men: role of visceral fat accumulation.

Diabetes Care 2004;27:2498-500.

21. Hunt KJ, Resendez RG, Williams K, et al. National Cholesterol Education Program

versus World Health Organization metabolic syndrome in relation to all-cause and

cardiovascular mortality in the San Antonio Heart Study. Circulation 2004;110:1251-

7.

22. Kern PA, Ranganathan S, Li C, et al. Adipose tissue tumor necrosis factor and

interleukin-6 expression in human obesity and insulin resistance. Am J Physiol

Endocrinol Metab 2001;280:E745-51.

23. Kerner A, Avizohar O, Sella R, Bartha P, Zinder O, Markiewicz W et al. Association

Between Elevated Liver Enzymes and C-Reactive Protein. Possible Hepatic Contribution

to Systemic Inflammation in the Metabolic Syndrome. Arterioscler Thromb Vasc Biol

2005;25:193-7.

24. Sakkinen P, Abbott RD, Curb JD, et al. C-reactive protein and myocardial infarction. J

Clin Epidemiol 2002;55:445-51.

25. Ridker PM, Hennekens CH, Buring JE, et al. C-reactive protein and other markers of

inflammation in the prediction of cardiovascular disease in women. N Engl J Med

2000;342:836-43.

26. Unger RH. Lipotoxic diseases. Annu Rev Med 2002;53:319-36.

27. Fishbein MH, Miner M, Mogren C, et al. The spectrum of fatty liver in obese children

and the relationship of serum aminotransferases to severity of steatosis. J Pediatr

Gastroenterol Nutr 2003;36:54-61.

28. Daniel S, Ben Menachem T, Vasudevan G, et al. Prospective evaluation of unexplained

chronic liver transaminase abnormalities in asymptomatic and symptomatic patients.

Am J Gastroenterol 1999;94:3010-4.

143

Chapter 9.1

ALANINE AMINOTRANSFERASE AND THE 6-YEAR RISK OF THE METABOLIC SYNDROME IN CAUCASIAN MEN AND WOMEN:

THE HOORN STUDY



Giel Nijpels2, 3

Lex M. Bouter2

Coen D. A. Stehouwer4

Robert J. Heine1,2

Michaela Diamant1

From the Departments of 1Endocrinology/Diabetes Center, 3General Practice, the

2EMGO Institute, VU University Medical Center, Amsterdam, The Netherlands, and

the 4Department of Internal Medicine, Academic Hospital Maastricht, Maastricht,

The Netherlands.

Diabetic Medicine 2007;24:430-5.

Reprinted with permission from Blackwell Publishing Ldt.

CHAPTER 9

144

ABSTRACT

Aims To study the association between alanine aminotransferase (ALT) and the 6-

year risk of the metabolic syndrome in a population-based study in Caucasian men

and women.

Methods The association of ALT with the 6-year risk of the metabolic syndrome in

1,097 subjects, aged 50-75 years, was assessed in the Hoorn Study with logistic

regression analysis. Subjects with the metabolic syndrome at baseline, defined

according to the Adult Treatment Panel III of the National Cholesterol Education

Program, were excluded.

Results After 6.4 (range: 4.4-8.1) years of follow-up, 226 subjects (20.6%) had

developed the metabolic syndrome. The odds ratio (95% confidence interval) for

developing the metabolic syndrome, adjusted for age, sex, alcohol-intake and

follow-up duration was 2.25 (1.50-3.37) for subjects in the upper tertile compared

to those in the lower tertile of ALT. This association persisted after additional

adjustment for all the baseline metabolic syndrome features [1.62 (1.02-2.58)].

Among the individual components of the metabolic syndrome, ALT was significantly

associated only with fasting plasma glucose at follow-up.

Conclusions These data suggest that ALT is associated with risk of the metabolic

syndrome in a general population of middle-aged Caucasian men and women,

further strengthening the role of ALT as an indicator for future metabolic

derangement. These findings warrant further studies to elucidate the role of non-

adipose tissue fat accumulation in the pathogenesis of metabolic syndrome related

complications.

ALANINE AMINOTRANSFERASE AND THE METABOLIC SYNDROME

145

INTRODUCTION

The metabolic syndrome, characterised by abdominal obesity, hyperglycaemia,

hypertension and dyslipidaemia, confers a high risk for the development of type 2

diabetes mellitus (DM2) [1,2] and cardiovascular disease (CVD) [3]. Individuals with

DM2 or the metabolic syndrome have a higher prevalence of elevated liver

enzymes, including alanine aminotransferase (ALT) [4,5]. Several cross-sectional

studies have demonstrated associations of ALT with individual components of the

metabolic syndrome, including obesity and dyslipidaemia [6,7].

To date, a limited number of studies have addressed the prospective relation of

liver enzymes with the future risk of the metabolic syndrome. Nakanishi et al found

that of the liver enzymes studied, ALT was associated with risk of metabolic

syndrome, but this study was limited to middle-aged Japanese men and used body

mass index (BMI) instead of waist circumference to define the metabolic syndrome

[8]. Hanley et al studied the relation of four different liver enzymes (including ALT)

with the development of the metabolic syndrome in a multi-ethnic cohort and

demonstrated that ALT was positively associated with risk of the metabolic

syndrome [9]. In a Japanese cohort of male and female health care employees,

Suzuki et al studies the sequential order of elevated transaminases (ALT and

aspartate aminotransferase) with feature of the metabolic syndrome, and found

that elevated transaminases are preceded by weight gain, while the other feature

are followed by elevated transaminases [10]. To date, no study has assessed the

prospective relation of ALT and the metabolic syndrome in a population-based

study in Caucasian men and women.

In the present study, we addressed the prospective association of ALT with the 6-

year risk of the metabolic syndrome, defined according to the Adult Treatment

Panel III of the National Cholesterol Education Program, and with the individual

components of the metabolic syndrome in a population-based study of glucose

tolerance and related complications in Caucasian men and women aged 50-75

years at baseline.

CHAPTER 9

146

MATERIALS AND METHODS

Study population and follow-up

The subjects were participants in the Hoorn Study, a prospective population-based

cohort study of glucose metabolism and diabetes complications [11]. In brief, in

1989, a random sample of all man and women aged 50 to75 was taken from the

municipal registry of the town of Hoorn in The Netherlands. Of the 3,552

individuals who where invited to take part in the study, 2,540 agreed to participate

(71%). Baseline data were collected from October 1989 through February 1992.

After excluding non-Caucasians (n=56) the study cohort consisted of 2,484 men

and women. Between January 1996 and December 1998, 2,086 participants were

invited to take part in a follow-up examination, the other members of the original

study population were not invited because of logistical reasons (n=140), or because

they died (n=150) or had moved out of Hoorn (n=108). A total of 1,513 subjects

(72.5%) took part in the follow-up measurements [12]. Metabolic syndrome at

baseline and at follow-up was defined according to the Adult Treatment Panel III of

the National Cholesterol Education Program, i.e. three or more of the following:

fasting glucose ≥ 6.1 mmol/l, HDL cholesterol <1.0 mmol/l (men) or <1.3 mmol/l

(women), triglycerides ≥ 1.7 mmol/l, waist-circumference ≥ 102 cm (men) or ≥ 88

cm (women) and hypertension ≥130/85 mmHg [13]. For the present study, subjects

with the metabolic syndrome and/or DM2 at baseline were excluded (n=329).

Furthermore subjects with missing data on alcohol intake, missing data on ALT

enzyme activity, insulin, smoking status and physical activity at baseline were also

excluded (n=22). Finally, subjects with missing data to define the metabolic

syndrome at follow-up (n=63) and subjects with ALT levels > 2.5 times of the upper

level of the reference value (n=2) were also excluded, the latter to reduce

confounding due to subjects with ALT elevation caused by viral or toxic agents,

resulting in 1,097 evaluable subjects. The Ethical Review Committee of VU

University Medical Centre approved the Hoorn Study and written informed consent

was obtained from all participants.

Measurements

Fasting blood samples were collected after an overnight fast from 20.00 hours the

evening before. Serum ALT enzyme activity was measured according to the 1985

method of the International Federation of Clinical Chemistry [14]. Glucose was

measured in plasma with the glucose dehydrogenase method (Merck, Darmstadt,


147

Germany) at the baseline measurement and with the hexokinase method

(Boehringer Mannheim, Germany) at the follow-up measurement. Immuno-specific

insulin was measured in serum with a double-antibody RIA (antibody SP21, Linco

Research, St. Louis, USA). Triglycerides, total and HDL-cholesterol were determined

in serum by enzymatic techniques (Boehringer Mannheim, Germany). Systolic and

diastolic blood pressure was measured twice on the right arm with a random-zero

sphygmomanometer (Hawksley-Gelman, Lancing, UK) and the average of both

measurements was used for the analyses. Information on alcohol intake was

assessed by a validated semi-quantitative food frequency questionnaire [15].

Subjects were asked whether they drink alcohol and if so, how many glasses of

alcoholic drinks they consumed in a week. This information was converted to

alcohol intake in grams/day with the computerised version of the Dutch Food

Composition Table. Physical activity was assessed as described previously [11].

Weight and height were measured in subjects wearing light clothes only, and the

BMI was calculated as weight divided by height squared (kg/m2). Waist and hip

circumference were measured according to a standardised method [16]. Smoking

status was assessed by questionnaire.


Data are presented as mean (SD) or as median (interquartile range) for variables

with a skewed distribution according to tertiles of ALT. For the prospective data,

logistic regression analyses were performed to calculate the odds ratios (ORs) and

the 95% confidence intervals (95% CIs) for the metabolic syndrome. The upper

tertile of ALT was compared with the lower tertiles of ALT. The first model was

adjusted for age, sex, alcohol-intake and follow-up duration, because follow-up

differed between subjects. In the final model, additional adjustments were made

for the individual components of the metabolic syndrome criteria entered as

continuous variables. To assess the prospective relation of ALT with the individual

components of the metabolic syndrome, we used multivariate linear regression

models. In these models the individual metabolic syndrome component at follow-

up was entered as the outcome variable and ALT as the determinant. These models

were adjusted for age, sex and follow-up-duration and subsequently adjusted for

the metabolic syndrome component at baseline. These associations were expressed

as standardised betas with the 95% CI. A standardised beta of 0.5 means that if the

independent variable increases by 1 SD, the dependent variable increases by 0.5

SD. Variables with a skewed distribution were entered in the models after

logarithmic transformation. The statistical analyses were performed with SPSS for

CHAPTER 9

148

Windows version 11.0.5. A two-sided P-value<0.05 was considered to indicate

statistical significance. Effect-modification by age and sex were tested by entering

interaction terms (age times ALT-tertile or sex times ALT-tertile) into the models.

We choose a two-sided P-value<0.1 to indicate effect-modification.

RESULTS

Baseline characteristics

The mean age of the 1,097 participants (465 men and 632 women) was 60.0 (SD

6.7) years. Table 9.1 shows the baseline characteristics of the participants stratified

by ALT divided into tertiles. Subjects in the upper tertile were younger, more likely

to be male, had higher fasting glucose and cholesterol levels, lower HDL-

cholesterol levels and a higher blood pressure, compared to those in the lower

tertile. A borderline significant trend was observed for fasting triglyceride levels.

Risk for development of the metabolic syndrome

Table 9.2 shows the odds ratio of development of the metabolic syndrome among

1,097 men and women who were free of the metabolic syndrome at baseline after a

mean of 6.4 (range: 4.4-8.1) years of follow-up with ALT divided into tertiles. Of

these, 226 (20.6%) subjects developed metabolic syndrome at follow-up. In the

model adjusted for age, sex and follow-up duration and alcohol-intake, subjects in

the upper tertile had an odds ratio of 2.25 (95%CI: 1.50-3.37) compared to those in

the lower tertile. The subsequent models in which components of the metabolic

syndrome were entered as continuous variables showed that this relation was most

strongly attenuated by waist, followed by glucose and to a lesser extent by blood

pressure, but the associations remained statistically significant (OR: 1.62 [1.02-

2.58]). Lifestyle factors, including physical activity and smoking, nor fasting insulin

affected the relation to a significant extent. No effect modification for sex or age

was observed (both P > 0.1). To further address the effect of alcohol-intake as a

major potential confounder, we excluded subjects (n=152) with an alcohol-intake of

more than 20 grams per day. However, this only lead to a small attenuation of

model 1 [2.10 (1.36-2.54)]. The models in which the components of the metabolic

syndrome were entered as dichotomised variables (according to the Adult

Treatment Panel III of the National Cholesterol Education Program) did not

materially change the associations presented in Model 1 (data not shown), the OR

of Model 2 was slightly higher (OR: 2.02 [1.31-3.12])


149

The analyses assessing the relation of ALT with the individual components of the

metabolic syndrome showed that ALT was associated with all the components at

follow-up except for HDL-cholesterol. After adjustment for the baseline values of

the individual components, ALT was significantly associated with glucose only

(Table 9.3).

Table 9.1. Baseline characteristics of the participants (n=1,097) stratified by ALT

tertiles

Variable Alanine aminotransferase activity (U/L)

Tertile 1 Tertile 2 Tertile 3 Ptrend*

N 329 418 350

ALT (U/L) 10.4 (2.1) 15.6 (1.54) 27.3 (10.1)

Age (years) 60.9 (6.9) 60.4 (6.8) 58.6 (6.3) <0.001

Male (%) 27.1 41.9 57.4 <0.001

Body mass index (kg/m2) 25.2 (2.9) 25.5 (2.7) 26.6 (3.0) <0.001

Waist (cm) 84.5 (9.5) 87.4 (8.7) 91.0 (8.8) <0.001

Fasting glucose (mmol/l) 5.2 (0.5) 5.3 (0.5) 5.4 (0.5) <0.001

Plasma lipids (mmol/l)

Total-cholesterol 6.5 (1.1) 6.5 (1.0) 6.6 (1.1) 0.016

HDL-cholesterol 1.42 (0.35) 1.41 (0.36) 1.38 (0.39) 0.020

Triglycerides 1.2 (0.9-1.5) 1.2 (0.9-1.5) 1.3 (1.0-1.6) 0.054

Systolic BP (mmHg) 127 (18) 130 (17) 131 (18) <0.001

Diastolic BP (mmHg) 78 (10) 79 (9) 82 (9) <0.001

Abbreviations: ALT, alanine aminotransferase; BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein Data are means (SD) or median (interquartile range) for variables with a skewed distribution; *Tested for trend with regression analyses adjusted for age and sex

CHAPTER 9

150

Table 9.2 The associations of ALT with development of the metabolic syndrome

(N=1,097)

ALT enzyme activity

Model Tertile 1 Tertile 2 Tertile 3

N 329 418 350

Model 1 1 1.43 (0.96-2.13) 2.25 (1.50-3.37)

Model 1 + waist 1 1.30 (0.89-1.95) 1.70 (1.11-2.58)

Model 1 + glucose 1 1.34 (0.89-2.01) 2.01 (1.33-3.03)

Model 1 + HDL-cholesterol 1 1.55 (1.03-2.32) 2.60 (1.72-3.95)

Model 1 + triglycerides 1 1.54 (1.03-2.35) 2.18 (1.43-3.33)

Model 1 + systolic BP 1 1.39 (0.93-2.08) 2.08 (1.38-3.14)

Model 1 + diastolic BP 1 1.41 (0.95-2.10) 2.11 (1.40-3.18)

Model 1 + life style factors 1 1.45 (0.97-2.18) 2.29 (1.51-3.48)

Model 1 + insulin 1 1.46 (0.97-2.20) 2.19 (1.44-3.35)

Model 2 1 1.40 (0.90-2.18) 1.62 (1.02-2.58)

*Model 1 adjusted for age, sex, alcohol-intake and follow-up duration; Model 2: Model 1 + adjusted for waist,

glucose, HDL-cholesterol, triglycerides and systolic and diastolic blood pressure

Life style factors: smoking and physical activity

Table 9.3. Multivariate linear regression analysis of ALT with the

individual metabolic syndrome components at follow-up*

Standardized betas (95%CI)

Metabolic syndrome

component at follow-up

Model 1 Model 2

Fasting glucose 0.19 (0.13-0.25) 0.13 (0.07;0.18)

Waist circumference 0.17 (0.11-0.22) 0.004 (-0.03;0.04)

Triglycerides 0.10 (0.03;0.16) 0.05 (-0.001;0.10)

Diastolic blood pressure 0.10 (0.04;0.17) 0.03 (-0.02;0.09)

Systolic blood pressure 0.09 (0.03;0;15) 0.03 (-0.02;0.08)

HDL-cholesterol 0.03 (-0.03;0.09) -0.02 (-0.05;0.02)

Models 1 and 2 adjusted for age, sex, alcohol-intake and follow-up duration,

Model 2 additionally adjusted for the baseline value of the individual component.

*Standardized betas indicate SD difference per 1 SD ALT at baseline


151

DISCUSSION

The main finding of this study is that in an elderly population of non-diabetic

Caucasian men and women free of the metabolic syndrome at baseline, ALT was

associated with an increased risk of conversion to the metabolic syndrome. This

association was independent of age, sex, alcohol-intake, follow-up duration and the

baseline metabolic syndrome components. In the prospective analyses of the

individual components of the metabolic syndrome, ALT was significantly associated

with fasting glucose levels and to a lesser extent with HDL-cholesterol, triglyceride

levels and elevated blood pressure and waist circumference.

The findings of the present study are in accordance with the two previous studies,

predominantly performed in Asian [8] and African American and Hispanic [9]

populations, that addressed the relation of liver enzymes including ALT with future

risk of the metabolic syndrome. Correction of all the baseline metabolic syndrome

criteria may result in over-correction, implying that the models in the present study

may underestimate the true associations. Other risk factors for the development of

the metabolic syndrome have been studied. These factors included C-reactive

protein (CRP) [17], physical inactivity [18], pro-insulin [19] and the individual

components themselves [19]. In the present study, physical activity did not affect

the associations. Unfortunately, pro-insulin and CRP levels were not available at

baseline in the Hoorn study cohort and consequently the effect of these mediating

or confounding variables could not be studied. The chronological ordering of

elevated transaminases (ALT and AST) in relation to components of the metabolic

syndrome was studied by Suzuki and et al, they reported, that weight gain, low

HDL-cholesterol and hypertrglyceridemia preceded transaminase elevation, follow

by hypertension and glucose intolerance [10]. Our study was not designed to study

the chronological ordering of ALT with the individual components of the metabolic

syndrome, however, our results do support some of the findings by Suzuki et al

that elevated ALT precedes components of the metabolic syndrome. Indeed,

glucose at follow-up adjusted for baseline glucose values was the strongest

component associated with elevated ALT. The mechanisms underlying the observed

association between ALT (and other liver enzymes) and the metabolic syndrome are

not yet fully understood. ALT is most strongly associated with liver fat

accumulation measured by proton-magnetic-resonance spectroscopy [20],

reflecting fat deposition in the liver, which is regarded as a feature of the metabolic

syndrome [21,22]. This hepatic fat deposition, as reflected by slightly elevated ALT

levels, may lead to early hepatic and systemic insulin resistance [23,24]. Indeed, an

CHAPTER 9

152

earlier study demonstrated that hepatic fat content measured by proton

spectroscopy is associated with insulin resistance, independent of obesity [25]. In

addition, hepatic fat content was related to a decreased suppression of insulin-

mediated hepatic glucose output, which is in line with the observation of the

present study that ALT is positively associated to fasting glucose levels in the

prospective analysis. High ALT levels might also reflect hepatic and systemic

inflammation, and the interplay of visceral adipose tissue, stimulating hepatic CRP

production which may contribute to further metabolic derangement. The results of

the present study show that subjects with slightly elevated ALT have a higher risk

of developing the metabolic syndrome. A number of potential limitations of this

study should be considered. The population consisted of Caucasian individuals

aged 50 to 75, and caution should be exercised to generalise our findings to other

populations. Although, we addressed alcohol as a confounder in our study, we

could not account for other factors that might have caused a rise in ALT, such as

viral hepatitis infections or hepatotoxic drugs. To overcome this limitation to some

extend, subjects with an ALT level of 2.5 times the upper level of the reference

value were excluded. In about on third of hepatitis C infections, a condition

associated with an increased risk for development of diabetes, subjects may have

normal ALT levels [26,27]. However, the effect of hepatitis C with regard to our

results may be very limited, as the prevalence of hepatitis C infection in the Dutch

general population is very low [28]. In addition, the association of other liver

enzymes such as asparatate aminotransferase and gamma-glutamyltransferase with

risk of developing the metabolic syndrome were not assessed. A previous study,

however, found ALT most strongly correlated to hepatic fat accumulation [20]. In

the Hoorn study, no diagnostic work-up was performed to assess the cause of

elevated ALT, however all relevant baseline data was send to the individual’s

general practitioner with an advice to perform a further diagnostic work-up

according to prevailing guidelines.

Despite the above mentioned limitations, the findings obtained in a population-

based study of elderly Caucasian men and women support the conclusion that ALT

is positively associated with the metabolic syndrome and further studies are

needed to elucidate and understand the pathophysiological mechanisms.


153

REFERENCES

1. Sattar N, Gaw A, Scherbakova O, et al. Metabolic syndrome with and without C-

reactive protein as a predictor of coronary heart disease and diabetes in the West of

Scotland Coronary Prevention Study. Circulation 2003;108:414-9.

2. Lorenzo C, Okoloise M, Williams K, Stern MP, Haffner SM. The metabolic syndrome as

predictor of type 2 diabetes: the San Antonio heart study. Diabetes Care 2003;

26:3153-9.

3. Dekker JM, Girman C, Rhodes T, et al. Metabolic syndrome and 10-year cardiovascular

disease risk in the Hoorn Study. Circulation 2005;112):666-73.


aminotransferase activity among Mexican Americans. Am J Epidemiol 1997;146:565-

71.


incident diabetes: the Mexico City diabetes study. Diabetes Care 2005;28:1757-62.




7. Salvaggio A, Periti M, Miano L, Tavanelli M, Marzorati D. Body mass index and liver

enzyme activity in serum. Clin Chem 1991;37:720-3.


metabolic syndrome and type 2 diabetes in middle-aged Japanese men. Diabetes Care

2004;27:1427-32.

9. Hanley AJ, Williams K, Festa A, et al. Liver markers and development of the metabolic

syndrome: the insulin resistance atherosclerosis study. Diabetes 2005;54:3140-7.

10. Suzuki A, Angulo P, Lymp J, et al. Chronological development of elevated

aminotransferases in a nonalcoholic population. Hepatology 2005;41:64-71.



1995;18:1270-3.

12. de Vegt F, Dekker JM, Jager A, et al. Relation of impaired fasting and postload glucose

with incident type 2 diabetes in a Dutch population: The Hoorn Study. JAMA

2001;285:2109-13.

13. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in

Adults. 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;285:2486-97.






CHAPTER 9

154

15. Grootenhuis PA, Westenbrink S, Sie CM, et al. A semiquantitative food frequency

questionnaire for use in epidemiologic research among the elderly: validation by

comparison with dietary history. J Clin Epidemiol 1995;48:859-68.

16. Mooy JM, Grootenhuis PA, de Vries H, et al. Intra-individual variation of glucose,

specific insulin and proinsulin concentrations measured by two oral glucose tolerance

tests in a general Caucasian population: the Hoorn Study. Diabetologia 1996;39:298-

305.

17. Han TS, Sattar N, Williams K, et al. Prospective study of C-reactive protein in relation

to the development of diabetes and metabolic syndrome in the Mexico City Diabetes

Study. Diabetes Care 2002;25:2016-21.

18. Laaksonen DE, Lakka HM, Salonen JT, et al. Low levels of leisure-time physical activity

and cardiorespiratory fitness predict development of the metabolic syndrome.


19. Palaniappan L, Carnethon MR, Wang Y, et al. Predictors of the incident metabolic

syndrome in adults: the Insulin Resistance Atherosclerosis Study. Diabetes Care

2004;27:788-93.




2004;47:1360-9.

21. Knobler H, Schattner A, Zhornicki T, et al. Fatty liver--an additional and treatable

feature of the insulin resistance syndrome. QJM 1999;92:73-9.

22. Garg A, Misra A. Hepatic steatosis, insulin resistance, and adipose tissue disorders. J

Clin Endocrinol Metab 2002;87:3019-22.




24. Bugianesi E, Gastaldelli A, Vanni E, et al. Insulin resistance in non-diabetic patients

with non-alcoholic fatty liver disease: sites and mechanisms. Diabetologia

2005;48:634-42.

25. Seppala-Lindroos A, Vehkavaara S, Hakkinen AM, et al. Fat accumulation in the liver is

associated with defects in insulin suppression of glucose production and serum free

fatty acids independent of obesity in normal men. J Clin Endocrinol Metab

2002;87:3023-8.

26. Noto H, Raskin P. Hepatitis C infection and diabetes. J Diabetes Complications

2006;20:113-20.

27. Bruce MG, Bruden D, McMahon BJ, et al. Hepatitis C infection in Alaska Natives with

persistently normal, persistently elevated or fluctuating alanine aminotransferase

levels. Liver International 2006;26:643-9.

28. Veldhuizen IK, Conyn-van Spaendonck MA, Dorigo-Zetsma JW. Seroprevalentie van

hepatitis B en C in de Nederlandse bevolking [Seroprevalence of hepatitis B and C in

the Dutch population]. Infectieziektenbulletin 1999;10:182-4.

155

Chapter 9.2

NO INDEPENDENT ASSOCIATION OF ALANINE AMINOTRANSFERASE WITH RISK OF FUTURE TYPE 2 DIABETES IN THE HOORN STUDY



Giel Nijpels2,3

Robert J. Heine1

Michaela Diamant1

From the Departments of 1Endocrinology/Diabetes Center and 3General Practice,

and the 2EMGO Institute, VU University Medical Center, Amsterdam, The

Netherlands

Published as letter in: Diabetes Care 2005;28:2812

Copyright © 2005 by the American Diabetes Association, Inc.

CHAPTER 9

156

We read with great interest the article by Nannipieri et al. [1], which reported no

association between aspartate aminotransferase, alanine aminotransferase (ALT), or

alkaline phosphatase and the 7-year incidence of impaired glucose tolerance and

type 2 diabetes. Interestingly, γ-glutamyltransferase was, but ALT was not, a

significant predictor of impaired glucose tolerance or type 2 diabetes. This is, as

the authors stated, in contrast to previously published papers that found significant

prospective associations of ALT with incident type 2 diabetes after adjustment for

obesity, insulin sensitivity, or inflammation (C-reactive protein) [2–4]. We studied

the baseline ALT enzyme activity in relation to 6-year incident type 2 diabetes in

the Hoorn Study, a population-based cohort study of glucose tolerance among

Caucasian men and women [5]. Glucose tolerance was assessed by oral glucose

tolerance test at baseline and after 6.4 years of follow-up. The study methods were

similar to the Mexico City Diabetes Study described by Nannipieri et al. [1], with the

exception that the participants of the Hoorn Study were 15 years older at baseline

(5). Of the 1,289 subjects free of type 2 diabetes at baseline, 123 developed type 2

diabetes during follow-up. We found that ALT was a significant predictor of type 2

diabetes only in the models adjusted for age, sex, and follow-up duration, with an

odds ratio of 2.18 (95% CI 1.29–3.69) for subjects in the upper tertile versus those

in the lower tertile. However, after additional adjustment for waist circumference,

BMI, alcohol consumption, fasting plasma insulin levels, and 2-h postload glucose

levels, the association attenuated and lost significance (1.18 [0.66–2.11]). The

other liver enzymes were not determined in the Hoorn Study. Thus, in accordance

with the report by Nannipieri et al. [1], but in contrast to the findings by others, we

found that ALT was not an independent predictor of incident type 2 diabetes in the

Hoorn study. We think it is important to report these findings, since we cannot

exclude the possibility of publication bias leading to overrepresentation of studies

showing an independent relationship between liver enzymes and risk of type 2

diabetes. The observed association between ALT and incident type 2 diabetes in

several published reports may be explained by the fact that these studies were

performed in selected populations, i.e., in high-risk populations and may not be

representative for the general population. An alternative explanation is that the

applied models overadjust for potential mediating variables such as insulin and 2-h

glucose, factors that might be directly related to liver fat. To increase the insight

into the role of liver fat in the pathophysiology of the metabolic syndrome and type

2 diabetes, further studies are needed to assess the underlying mechanisms. The

role of oxidative stress, as hypothesized by Nannipieri et al. [1], as well as the

contribution of inflammation should be explored in large-scaled prospective

studies.

ALANINE AMINOTRANSFERASE AND INCIDENT TYPE 2 DIABETES MELLITUS

157

REFERENCES


incident diabetes: the Mexico City Diabetes Study. Diabetes Care 2005;28:1757–62.



diabetes. Diabetes 2002;51:1889–95.


type 2 diabetes: the Insulin Resistance Atherosclerosis Study. Diabetes

2004;53:2623–32.



syndrome, and C-reactive protein in the West of Scotland Coronary Prevention Study.

Diabetes 2004;53:2855–60.


intolerance in a Dutch Caucasian population: the Hoorn Study. Diabetes Care

1995;18:1270–3.

CHAPTER 9

158

159

Chapter 10.1

GENERAL DISCUSSION - METHODOLOGICAL CONSIDERATIONS

CHAPTER 10

160

STUDY DESIGN AND STUDY POPULATIONS

In this thesis both cross-sectional and prospective studies were used to assess the

association of determinants with outcome measures. In cross-sectional studies, the

determinants of interest and one or more outcome variables are measured at the

same time. Thus, causal inference remains hypothetical, but nevertheless, these

studies may contribute to a better insight in pathogenic processes. A causal

relationship implies that the putative cause precedes the effect. In a prospective

study design, determinants are measured before onset of the outcome, allowing

assessment of the temporal relationship between determinant and outcome

measure. However, merely a temporal relationship does not necessarily imply a

causal relationship between the studied variables. More requirements, as proposed

by Hill, need to be fulfilled to establish a true causal relationship between exposure

or determinant and the outcome [1]. Although neither cross-sectional nor

prospective studies can provide formal prove for a causal relation, the prospective

study design is generally considered to be more informative.

In Chapters 2 to 5 (determinants of postprandial metabolic changes and responses

in relation to cardiovascular disease [CVD] risk) and Chapter 7 (ALT in relation to

endothelial dysfunction and insulin sensitivity), cross-sectional study designs were

used. In Chapters 8 and 9, a prospective study design was used to assess the

relation of ALT as the determinant of interest with outcome. As pointed out before,

a significant prospective association does not necessarily imply causality, but

provides a stronger insight in the possible pathogenic mechanisms.

Selection bias may affect the internal validity of the study, such that a distortion of

the estimate of the association between determinant and outcome can occur,

resulting in different estimates for participants and non-participants. In Chapters 2

to 5, data from the Hoorn Prandial study was used [2]. The relative paucity of data

in women and the disproportionately high relative risk of CVD in especially post-

menopausal women [3,4], prompted us to perform the study in women only.

Moreover, postmenopausal women have higher postprandial triglyceride

concentrations compared to pre-menopausal women [5]. The participants of the

Hoorn Prandial study were randomly selected from the municipal registry of the city

of Hoorn and from the registry from the Diabetes Care System in Hoorn. In these

studies selection bias may indeed have occurred, since only 24% of the invited

individuals were willing to participate in the study [2], possibly resulting in a

relatively healthy selection of women with DM2 and controls. On the other hand,

GENERAL DISCUSSION

161

the possibility of participation of relatively unhealthy control subjects (participating

in the study as a health check-up) cannot be ruled out. In Chapter 7, the

participants were recruited from shared care projects in Amsterdam and Hoorn,

and through newspaper advertisements. Those who participated were willing to

undergo the extensive study protocol. This may have resulted in a selection of

participants who were relatively healthier than those who were unwilling to

participate (volunteer bias). For Chapters 8 and 9, data from the Hoorn Study were

used. Participants in this study were randomly selected from the municipal registry

from the city of Hoorn and a relatively large number of the invited persons

participated in the baseline (71.5% of those invited) and follow-up examinations

(72.5% of those invited) [6,7], this may considered to be a representative sample of

the general population. In Chapter 8, we excluded individuals with missing

information on non-fatal disease because they did not give permission to access

their hospital records. We have studied the possibility of selection bias and

repeated the analyses for all-cause mortality using the data from the entire original

Hoorn study cohort, without prevalent CVD and coronary heart disease, which

yielded almost identical estimates.

DETERMINANTS OF INTEREST

In Chapter 2, we studied the associations of clinical and biochemical variables with

fasting and postprandial glucose and triglycerides. The clinical and biochemical

variables included anthropometric measures, laboratory analyses and information

on physical activity and alcohol-intake assessed by questionnaires. Anthropometric

measurements were determined in duplicate according to a predefined protocol.

Medical history, medication, (former) smoking and alcohol use were assessed by

questionnaires [8] and physical activity was assessed by the Short Questionnaire to

Assess Health-enhancing physical activity of which reproducibility and relative

validity were described previously [9]. The inter-assay coefficients of variation of

the laboratory analyses were <7%.

In Chapters 3 to 5, we measured fasting and postprandial glucose and triglyceride

responses following two consecutive fat-rich or carbohydrate-rich meals and

expressed these postprandial changes as incremental and/or total area under the

curve [10]. Most postprandial studies applied a single fat-rich or carbohydrate-rich

meal (e.g. [11-14]). However, glucose excursions following a first meal have been

shown to influence the responses after a second meal [15], and type and amount of

carbohydrate at breakfast have been shown to influence glucose, insulin and free

CHAPTER 10

162

fatty acid (FFA) responses to a subsequent lunch [16]. We assumed that the test-

meal induced responses reflect “real-life” postprandial metabolic changes that have

caused or are in part responsible for clinical or biochemical changes, i.e. structural

and functional vascular changes in the past. However, it is important to notice that

these postprandial responses may have been affected by life-style and/or

medication interventions. This might be especially true in patients with DM2. In

addition, these interventions may also have affected outcome measures. We applied

two consecutive meals with different meal composition. The carbohydrate-rich meal

consisted of 4 g fat, 162 g carbohydrates and 22 g of proteins and the fat-rich meal

consisted of 50 g fat, 56 g carbohydrates and 28 g proteins. The fat-rich meal

should be considered as a mixed meal because it contained carbohydrates, through

which beta-cell stimulation and insulin secretion will be induced, which may have

resulted in a lower postprandial triglyceride response [17], and possibly an

underestimation of the presented associations.

In Chapters 7 to 9, we used the enzyme activity of the liver enzyme alanine

aminotransferase (ALT) as a marker for non-alcoholic fatty liver disease (NAFLD).

The use of surrogate markers has been widely adopted in clinical and

epidemiological research [18], and a number of them have been validated against

the so-called “gold standard”, a diagnostic test, accepted as the most valid test to

diagnose the disease or condition of interest. In the case of NAFLD, direct

measurement of hepatic fat by means of liver biopsies, is regarded as the “gold

standard” [19,20]. However, in epidemiological and clinical studies the use of this

technique is limited, because of the elaborate and invasive nature and associated

risks. Non-invasive imaging techniques, such as ultrasound, computed tomography

and proton magnetic resonance spectroscopy (1H-MRS), have been used instead of

biopsies [21,22]. Ultrasound examination of the liver can be used to indicate the

presence or absence of fatty infiltration of the liver. However, at present, this

technique at its best can provide only semi-quantitative data with regard to NAFLD. 1H-MRS has been validated against direct determination of triglyceride content in

human liver biopsies [23], and may be used for clinical purposes, but has limited

applicability in larger populations, because of logistics and high costs. In

epidemiological studies, ALT has been used as a surrogate marker for liver fat

accumulation and to indicate the presence of NAFLD [24]. Westerbacka and

colleagues assessed the correlation of ALT with hepatic triglyceride content by 1H-

MRS and found a modest, but significant correlation (r=0.5) [25]. In most studies in

patients with biopsy proven NAFLD, ALT was higher compared to healthy controls,

although one study reported on NAFLD patients with “normal” ALT levels (i.e. below

GENERAL DISCUSSION

163

the laboratory reference range) [26]. Studies assessing the specificity and

sensitivity of ALT as a marker of NAFLD are limited. The sensitivity and specificity

of ALT in the diagnosis of NAFLD are low and depend on the cut-off value used

[27]. Therefore, lower cut-off values than the current upper limit of the laboratory

reference range have been proposed [28], but, since cut-off values depend on the

assay used, this proposal has been questioned by others [29,30]. In the studies

presented in this thesis, we chose not to define a cut-off value to delineate the

presence or absence of NAFLD. Instead we modeled ALT as a continuous variable or

divided into tertiles. In order to avoid confounding by other factors that might

cause ALT elevation, a detailed history on alcohol intake is mandatory to exclude

those patients with ALT elevation caused by alcohol-abuse, and to address alcohol-

intake as a possible confounder or effect-modifier. With these limitations in mind,

we considered ALT to be an acceptable marker for NAFLD in epidemiological

studies. However, we cannot rule out the possibility of non-differential

misclassification, as ALT is not always elevated in patients with NAFLD.

Furthermore, ALT may be elevated due to causes other than NAFLD. Unfortunately,

no information on viral and autoimmune liver disease and hepatotoxic medication

that might have caused ALT elevation in the presented studies was available, and

therefore we cannot rule out the possibility of non-differential misclassification.

Because non-differential misclassification may lead to an underestimation of the

true estimate, our reported estimates may be regarded as conservative. As we [24]

and others [31] have argued, measurement of ALT enzyme activity in fresh samples

(non-frozen) is of utmost importance to avoid loss of enzyme activity after repeated

freezing and thawing, which may further contribute to information bias (non-

differential misclassification). In the studies presented in this thesis in which ALT

enzyme activity was assessed, measurements were performed in previously non-

frozen blood samples. The interassay coefficient of variation of the ALT assay was

2.3%.

OUTCOME VARIABLES

Fasting and postprandial triglyceride and glucose concentrations after two

consecutive fat-rich and carbohydrate-rich meals were used as outcome variables in

Chapter 2 in order to study a number of determinants of these responses.

Triglycerides and glucose were assessed by precise and validated methods, with

interassay coefficients of variation of 2.2% and 1.3%, respectively. In Chapter 3, the

intima-media thickness of the carotid artery (cIMT) was used as outcome variable. It

is a well-established and widely applied marker of sub-clinical atherosclerosis in

CHAPTER 10

164

epidemiological studies [32]. cIMT was measured by a single observer with an intra-

observer coefficient of variation of 5.2%. Markers of fasting and postprandial

glycoxidative and lipoxidative stress were assessed in Chapter 4. Oxidized LDL was

measured with a validated commercially available competitive ELISA with inter-assay

and intra-assay coefficients of variation of 7.8% and 4.8%, respectively. Unbound

CML and CEL were measured in EDTA-plasma by high-performance liquid

chromatography with stable-isotope-dilution tandem mass spectrometric detection

(LC-MS/MS) using a procedure developed for the analysis of protein-bound CML and

CEL [33], with intra-assay and inter-assay coefficients of variation for CML of 3% and

4%, respectively and for CEL 5% and 9%, respectively. 3DG was measured with a

newly developed LC-MS/MS method after pre-column derivatization according to a

published procedure [34]. Intra-assay and inter-assay coefficients of variation for

3DG were 8% and 12%, respectively. Myeloperoxidase was determined in duplicate

by a commercially available sandwich ELISA in EDTA-plasma. The intra-assay and

inter-assay coefficients of variation were 3.3% and 5.0%, respectively (Chapter 5). In

Chapter 7, brachial artery, endothelium-dependent flow-mediated vasodilatation

(FMD) and whole body insulin sensitivity (M/I-value) as assessed with the

euglycemic hyperinsulinemic clamp technique, were used as outcome variables.

Important limitations for the FMD technique are the need for ultrasonographic

expertise and the significant day-to-day [35,36], intra and inter-observer, and

diurnal variability [37]. The intra-observer coefficient of variation of the FMD

measurements in Chapter 7 was 10.1% [38]. We used the euglycemic

hyperinsulinemic clamp technique, which is regarded as the “gold standard” for

measuring insulin sensitivity. However, this technique provides information on

whole body insulin sensitivity instead of the more relevant hepatic insulin

sensitivity with respect to NAFLD.

CONFOUNDING AND EFFECT MODIFICATION

In epidemiological studies it is common to adjust for putative confounding

variables and to present the estimates both before and after adjustment. A

confounder is related to both exposure and outcome, and should not be involved

as an intermediate variable. However, how to select potential confounders? This is

still debated. This can be illustrated in Chapter 8.2 in which we could not

demonstrate an independent association of ALT with incident DM2. We argued that

the applied models might have over-adjusted for potential mediating variables such

as insulin and 2-hour glucose. However, due to uncertainty of the pathogenic

GENERAL DISCUSSION

165

processes involved, a clear distinction between confounding and mediating

variables is not always possible. In addition, unmeasured confounding variables

may also play an important role. In our studies in which ALT was used as a

determinant, we could not rule out the effect of other causes besides alcohol that

might have resulted in elevated ALT activity. However, for example, the prevalence

of viral infections in the general population is relatively low [39], and therefore

probably did not contribute to bias of the estimates. In the studies presented in

this thesis, the analyses were adjusted for the appropriate confounding variables.

Effect modification is present if the magnitude of an association of a determinant

and an outcome is not uniform across a third variable. The higher relative CVD risk

and the paucity of studies in postmenopausal women, prompted us to perform the

postprandial studies (Chapter 2 to 5) in postmenopausal women only, which limits

the external validity, but yields a relatively homogeneous group. We presented the

analyses separately for women with NGM and DM2 to account for possible effect-

modification by DM2-status. In addition, we considered women with DM2 who used

statin medication as a separate study group, as statin medication is known to

influence lipid metabolism [40] and cIMT [41]. In Chapters 7 to 9 we addressed the

possibility of effect-modification by sex and when applicable by DM2-status.

REFERENCES

1. HILL AB. The environment and disease: association or causation? Proc R Soc Med

1965;58:295-300.

2. Alssema M, Schindhelm RK, Dekker JM, et al. Postprandial glucose and not

triglyceride concentrations are associated with carotid intima media thickness in

women with normal glucose tolerance: the Hoorn prandial study. Atherosclerosis (in

press).

3. Kaseta JR, Skafar DF, Ram JL, et al. Cardiovascular disease in the diabetic woman. J

Clin Endocrinol Metab 1999;84:1835-8.

4. Lewis SJ. Cardiovascular disease in postmenopausal women: myths and reality. Am J

Cardiol 2002;89:5E-10E.



1999;19:2737-41.

6. Mooy JM, Grootenhuis PA, de Vries H, et al. Intra-individual variation of glucose,

specific insulin and proinsulin concentrations measured by two oral glucose tolerance

tests in a general Caucasian population: the Hoorn Study. Diabetologia 1996;39:298-

305.

CHAPTER 10

166



2001;285:2109-13.



1995;18:1270-3.



Epidemiol 2003;56:1163-9.

10. Allison DB, Paultre F, Maggio C, et al. The use of areas under curves in diabetes

research. Diabetes Care 1995;18:245-50.





12. Vogel RA, Corretti MC, Plotnick GD. Effect of a single high-fat meal on endothelial

function in healthy subjects. Am J Cardiol 1997;79:350-4.

13. Anderson RA, Evans ML, Ellis GR, et al. The relationships between post-prandial

lipaemia, endothelial function and oxidative stress in healthy individuals and patients

with type 2 diabetes. Atherosclerosis 2001;154:475-83.

14. Ong PJ, Dean TS, Hayward CS, et al. Effect of fat and carbohydrate consumption on

endothelial function. Lancet 1999;354:2134.

15. Mari A, Tura A, Gastaldelli A, Ferrannini E. Assessing insulin secretion by modeling in

multiple-meal tests: role of potentiation. Diabetes 2002;51 Suppl 1:S221-S226.

16. Wolever TM, Bentum-Williams A, Jenkins DJ. Physiological modulation of plasma free

fatty acid concentrations by diet. Metabolic implications in nondiabetic subjects.


17. Kriketos AD, Sam W, Schubert T, et al. Postprandial triglycerides in response to high

fat: role of dietary carbohydrate. Eur J Clin Invest 2003;33:383-9.

18. Kluft C. Principles of use of surrogate markers and endpoints. Maturitas

2004;47:293-8.

19. Brunt EM, Janney CG, Di Bisceglie AM, et al. Nonalcoholic steatohepatitis: a proposal

for grading and staging the histological lesions. Am J Gastroenterol 1999;94:2467-

74.


2001;21:3-16.

21. Saverymuttu SH, Joseph AE, Maxwell JD. Ultrasound scanning in the detection of

hepatic fibrosis and steatosis. Br Med J (Clin Res Ed) 1986;292:13-5.

22. Ricci C, Longo R, Gioulis E, et al. Noninvasive in vivo quantitative assessment of fat

content in human liver. J Hepatol 1997;27:108-13.

GENERAL DISCUSSION

167

23. Szczepaniak LS, Babcock EE, Schick F, et al. Measurement of intracellular triglyceride

stores by H spectroscopy: validation in vivo. Am J Physiol 1999;276:E977-E989.

24. Schindhelm RK, Diamant M, Dekker JM, et al. Alanine aminotransferase as a marker of

non-alcoholic fatty liver disease in relation to type 2 diabetes mellitus and

cardiovascular disease. Diabetes Metab Res Rev 2006;22:437-43.




2004;47:1360-9.

26. Mofrad P, Contos MJ, Haque M, et al. Clinical and histologic spectrum of nonalcoholic

fatty liver disease associated with normal ALT values. Hepatology 2003;37:1286-92.

27. Zelber-Sagi S, Nitzan-Kaluski D, Halpern Z, Oren R. Prevalence of primary non-

alcoholic fatty liver disease in a population-based study and its association with

biochemical and anthropometric measures. Liver Int 2006;26:856-63.

28. Prati D, Taioli E, Zanella A, et al. Updated definitions of healthy ranges for serum

alanine aminotransferase levels. Ann Intern Med 2002;137:1-10.


2002;137:49-51.



31. Clark JM, Diehl AM. Defining nonalcoholic fatty liver disease: implications for

epidemiologic studies. Gastroenterology 2003;124:248-50.

32. Poredos P. Intima-media thickness: indicator of cardiovascular risk and measure of

the extent of atherosclerosis. Vasc Med 2004;9:46-54.

33. Teerlink T, Barto R, Ten Brink HJ, Schalkwijk CG. Measurement of Nepsilon-

(carboxymethyl)lysine and Nepsilon-(carboxyethyl)lysine in human plasma protein by

stable-isotope-dilution tandem mass spectrometry. Clin Chem 2004;50:1222-8.

34. Zwiener C, Glauner T, Frimmel FH. Method optimization for the determination of

carbonyl compounds in disinfected water by DNPH derivatization and LC-ESI-MS-MS.

Anal Bioanal Chem 2002;372:615-21.

35. Hijmering ML, Stroes ES, Pasterkamp G, et al. Variability of flow mediated dilation:

consequences for clinical application. Atherosclerosis 2001;157:369-73.

36. de Roos NM, Bots ML, Schouten EG, Katan MB. Within-subject variability of flow-

mediated vasodilation of the brachial artery in healthy men and women: implications

for experimental studies. Ultrasound Med Biol 2003;29:401-6.

37. Ringqvist A, Caidahl K, Petersson AS, Wennmalm A. Diurnal variation of flow-mediated

vasodilation in healthy premenopausal women. Am J Physiol Heart Circ Physiol

2000;279:H2720-H2725.

38. Van Dijk RA. Large artery properties in metebolic disease; studies in individuals wirh

type 2 diabetes, impaired glucose tolerance and hyperhomocyteinaemia [thesis].

Amsterdam: Vrije Universiteit, 2002.

CHAPTER 10

168

39. De Melker HE, Conyn-van Spaendonck MA. Immunosurveillance and the evaluation of

national immunization programmes: a population-based approach. Epidemiol Infect

1998;121:637-43.

40. Ginsberg HN. REVIEW: Efficacy and Mechanisms of Action of Statins in the Treatment

of Diabetic Dyslipidemia. J Clin Endocrinol Metab 2006;91:383-92.

41. Kang S, Wu Y, Li X. Effects of statin therapy on the progression of carotid

atherosclerosis: a systematic review and meta-analysis. Atherosclerosis

2004;177:433-42.

169

Chapter 10.2 – Part 1

GENERAL DISCUSSION - PATHOPHYSIOLOGICAL CONSIDERATIONS

POSTPRANDIAL DYSMETABOLISM, CARDIOVASCULAR DISEASE RISK AND TYPE 2 DIABETES MELLITUS

CHAPTER 10

170

INTRODUCTION

Already in 1979, Zilversmit postulated that atherogenesis might be a postprandial

phenomenon [1]. Indeed, a number of studies suggested that postprandial or post-

load glucose and postprandial triglyceride concentrations are much stronger

associated with some CVD risk factors and mortality than the respective fasting

values [2-8]. There is, however, still debate as to whether postprandial glucose is

causally related to CVD risk factors or merely a risk marker [9,10], i.e. reflecting the

underlying metabolic abnormalities in for example lipid metabolism. To date, most

studies used a single, often artificially composed liquid fat and/or carbohydrate

load to assess postprandial responses. However, in normal daily life, the metabolic

alterations following breakfast will affect the responses after lunch, thus potentially

resulting in a cumulative effect of breakfast on the plasma glucose and

triglycerides following lunch [11], especially in insulin resistant patients with type 2

diabetes mellitus (DM2). The main objective of the first part of this thesis was to

assess the relative contributions and elaborate on the potential mechanisms of

postprandial glucose and postprandial triglyceride levels following two consecutive

(breakfast and lunch) fat-rich or carbohydrate-rich meals to CVD risk in

postmenopausal women with DM2 and in women with normal glucose metabolism

(NGM).

DETERMINANTS OF POSTPRANDIAL GLUCOSE AND TRIGLYCERIDES

In Chapter 2, we have shown that the determinants of postprandial glucose and

triglyceride responses differ. Fasting triglycerides were the main determinant of

postprandial triglycerides, whereas postprandial glucose was determined by other

factors than fasting glucose values. Although women with DM2 had higher total

postprandial triglyceride responses (total area under the curve) as compared to

women with NGM, we found no significant differences in postprandial increase in

triglyceride concentrations (incremental area under the curve) between DM2 and

NGM. We expected a greater increase in postprandial triglyceride concentrations in

women with DM2, especially following the second meal (lunch). The relatively low

triglyceride response in women with DM2 might be the result of the meal-

composition. The applied fat-rich meal should be considered as a mixed-meal,

because of the substantial amount of carbohydrates. The resulting insulin response

suppresses the VLDL-secretion by the liver, which in turn attenuates the

postprandial triglyceride response. Furthermore, the women with DM2 included in

GENERAL DISCUSSION

171

our study were in relatively good glycemic control, possibly contributing to an

ameliorated response of the liver to the meal-induced insulin response. Finally, we

measured up to 8 hours after the first meal was given (4 hours after the second

meal). It may be conceivable that differences in postprandial triglyceride

metabolism become evident only after a longer study period (i.e. 24 hours), due to

differences in clearance between NGM and DM2 [12]. These explanations are

speculative, as we did not measure the triglyceride kinetics in our studies.

POSTPRANDIAL GLUCOSE AND TRIGLYCERIDE RESPONSES IN

RELATION TO CVD RISK

In the Hoorn Prandial study we addressed the question whether postprandial

glucose or postprandial triglyceride concentrations were more closely related to

CVD risk. In Chapter 3, we showed that postprandial glucose, and not the

triglyceride concentrations, are associated with carotid intima-media thickness

(cIMT) in women with NGM. The association between postprandial glucose and cIMT

was found for the meals, irrespective of their composition. We also showed that

measuring glucose 2 hours after the first meal was enough to assess the

association between postprandial glucose and cIMT. In contrast, plasma glucose

measured two hours after a 75-g oral glucose tolerance test (OGTT), was not as

strongly associated with cIMT in women with NGM. This could not be attributed to

limited precision of the single glucose measurement after the OGTT, because we

found that a single 2-hour glucose measurement following a meal, although

borderline significant, also was associated with cIMT. Based on this observation we

concluded that postprandial glucose concentrations following a meal may provide

additional clinical information above and beyond the 2-hour post-load glucose

following an OGTT.

In women with DM2, an association of cIMT was found with fasting glucose but not

with postprandial glucose concentrations. A possible explanation for the lack of an

association of postprandial glucose concentrations with cIMT in DM2, may be the

fact that women with DM2 were treated with blood glucose lowering medication.

These variably affect postprandial glucose, dependent on the mode of action of the

agents. In addition, some blood glucose lowering agents and life-style interventions

have been shown to retard IMT progression [13-15].

CHAPTER 10

172

Based on previously published papers, we would have expected an association

between postprandial triglycerides and cIMT, not only in patients with DM2 [3,7],

but also in healthy subjects [8,16]. However, no such associations between were

found in the study described in Chapter 3. In the Atherosclerosis Risk in

Communities (ARIC) Study, postprandial triglyceride levels were associated with

early atherosclerosis, but only in non-obese subjects [16]. The ARIC investigators

suggested that this association might be explained by high levels of atherogenic

triglyceride-rich lipoprotein remnants in the postprandial state. These lipoprotein

remnants were indeed related to cIMT, independent of postprandial triglycerides

[17]. Since we did not measure lipoprotein remnant particles in the present study,

the question regarding the presence or absence of an association between

postprandial lipids and cIMT in our study cannot be conclusively answered. It

should furthermore be considered that gender-specific studies on the relation

between postprandial triglycerides and CVD risk are scarce. The above-mentioned

ARIC Study reported an association between postprandial triglycerides and cIMT in

both men and women [16], whereas in another study the association was only

present in men and not in women [18].

MECHANISMS LINKING POSTPRANDIAL DYSMETABOLISM TO CVD

Glycoxidative and lipoxidative stress, AGE-formation and inflammation are

implicated as potential mechanisms that may link postprandial dysmetabolism to

CVD risk [19-22]. In chapter 4, we studied acute changes in markers of

glycoxidative and lipoxidative stress, including oxidized low-density-lipoprotein

(oxLDL), Nε-(carboxyethyl)-lysine (CEL), Nε-(carboxymethyl)-lysine (CML) and 3-

deoxyglucosone (3DG), following two consecutive fat-rich or carbohydrate-rich

meals. We found that oxLDL, 3DG and CML increased in the postprandial state. The

fasting concentrations of 3DG, oxLDL and CML in women with DM2 were

comparable with the postprandial concentrations of these markers in women with

NGM (Chapter 4). These finding extend a previous study from our group,

describing an increase in oxLDL and malondialdehyde in healthy young men

following 2 consecutive fat-rich mixed meals [23]. The increased postprandial

concentrations of oxLDL and 3DG may contribute to functional and structural

vascular changes which may contribute to the increased CVD risk. Indeed, the

postprandial increase in oxidative stress was associated with an impaired

postprandial flow-mediated vasodilatation in healthy young men in the

aforementioned study [23]. Future studies should address the relation of

GENERAL DISCUSSION

173

postprandial oxidative stress markers with vascular (dys)function in patients with

DM2.

Postprandial dysmetabolism may affect leukocyte recruitment and/or activation. In

turn, myeloperoxidase (MPO), which is expressed in leukocytes and released upon

activation, has been associated with CVD [24-26]. In Chapter 5, we hypothesized

that MPO would increase in the postprandial state, due to postprandial leukocyte

recruitment and/or activation, especially in individuals with DM2. However, in

contrast to our hypothesis, MPO decreased in NGM (both meal types) and in DM2

(fat-rich meals). Moreover, despite an increase of leukocytes following the fat-rich

meals in both NGM and DM2, no postprandial MPO increase was observed. The

mechanism by which a postprandial increase in number of leukocytes contributes

to CVD risk remains unclear. One study has reported that the postprandial increase

in number of leukocytes was associated with an impaired flow-mediated

vasodilatation [27].

CONCLUSIONS

The association of postprandial glucose with cIMT in women with NGM was

stronger than that of postprandial triglycerides. The association between

postprandial glucose and cIMT in women with NGM suggests that postprandial

glucose, in the normal range, is a marker or a risk factor of CVD. Possible

mechanisms that may contribute to the association between postprandial glucose

and CVD risk may include generation of postprandial oxidative stress. Future

studies should address the widely debated question whether lowering postprandial

glucose and oxidative stress may lower CVD risk.

REFERENCES 1. Zilversmit DB. Atherogenesis: a postprandial phenomenon. Circulation 1979;60:473-

85.

2. Hanefeld M, Koehler C, Henkel E, et al. Post-challenge hyperglycaemia relates more

strongly than fasting hyperglycaemia with carotid intima-media thickness: the RIAD

Study. Risk Factors in Impaired Glucose Tolerance for Atherosclerosis and Diabetes.

Diabet Med 2000;17:835-40.

CHAPTER 10

174



2000;23:1401-6.

4. Temelkova-Kurktschiev TS, Koehler C, Henkel E, et al. Postchallenge plasma glucose

and glycemic spikes are more strongly associated with atherosclerosis than fasting

glucose or HbA1c level. Diabetes Care 2000;23:1830-4.



2001;285:2109-13.

6. Chen X, Tian H, Liu R. Association between fasting and postprandial triglyceride

levels and carotid intima-media thickness in type 2 diabetes patients. Chin Med J

(Engl ) 2003;116:1933-5.

7. Ahmad J, Hameed B, Das G, et al. Postprandial hypertriglyceridemia and carotid

intima-media thickness in north Indian type 2 diabetic subjects. Diabetes Res Clin

Pract 2005;69:142-50.

8. Boquist S, Ruotolo G, Tang R, et al. Alimentary lipemia, postprandial triglyceride-rich

lipoproteins, and common carotid intima-media thickness in healthy, middle-aged

men. Circulation 1999;100:723-8.



10. Heine RJ, Balkau B, Ceriello A, et al. What does postprandial hyperglycaemia mean?

Diabet Med 2004;21:208-13.



Eur J Nutr 2005;44:377-83.

12. Dane-Stewart CA, Watts GF, Barrett PH, et al. Chylomicron remnant metabolism

studied with a new breath test in postmenopausal women with and without type 2

diabetes mellitus. Clin Endocrinol (Oxf) 2003;58:415-20.

13. Hanefeld M, Chiasson JL, Koehler C, et al. Acarbose slows progression of intima-

media thickness of the carotid arteries in subjects with impaired glucose tolerance.

Stroke 2004;35:1073-8.

14. Katakami N, Yamasaki Y, Hayaishi-Okano R, et al. Metformin or gliclazide, rather than

glibenclamide, attenuate progression of carotid intima-media thickness in subjects

with type 2 diabetes. Diabetologia 2004;47:1906-13.

15. Kim SH, Lee SJ, Kang ES, et al. Effects of lifestyle modification on metabolic

parameters and carotid intima-media thickness in patients with type 2 diabetes

mellitus. Metabolism 2006;55:1053-9.

16. Sharrett AR, Chambless LE, Heiss G, et al. Association of postprandial triglyceride and

retinyl palmitate responses with asymptomatic carotid artery atherosclerosis in

middle-aged men and women. The Atherosclerosis Risk in Communities (ARIC) Study.


GENERAL DISCUSSION

175

17. Karpe F, Boquist S, Tang R, et al. Remnant lipoproteins are related to intima-media

thickness of the carotid artery independently of LDL cholesterol and plasma

triglycerides. J Lipid Res 2001;42:17-21.

18. Tentor J, Harada LM, Nakamura RT, et al. Sex-dependent variables in the modulation

of postalimentary lipemia. Nutrition 2006;22:9-15.

19. Brownlee M. Glycation products and the pathogenesis of diabetic complications.


20. Brownlee M. The pathobiology of diabetic complications: a unifying mechanism.

Diabetes 2005;54:1615-25.

21. Baynes JW, Thorpe SR. Role of oxidative stress in diabetic complications: a new

perspective on an old paradigm. Diabetes 1999;48:1-9.

22. Bavenholm PN, Efendic S. Postprandial hyperglycaemia and vascular damage--the

benefits of acarbose. Diab Vasc Dis Res 2006;3:72-9.




24. Baldus S, Heeschen C, Meinertz T, et al. Myeloperoxidase serum levels predict risk in

patients with acute coronary syndromes. Circulation 2003;108:1440-5.

25. Exner M, Minar E, Mlekusch W, et al. Myeloperoxidase predicts progression of carotid

stenosis in states of low high-density lipoprotein cholesterol. J Am Coll Cardiol

2006;47:2212-8.

26. Zhang R, Brennan ML, Fu X, et al. Association between myeloperoxidase levels and

risk of coronary artery disease. JAMA 2001;286:2136-42.

27. Van Oostrom AJ, Sijmonsma TP, Verseyden C, et al. Postprandial recruitment of


CHAPTER 10

176

177

Chapter 10.2 – Part 2

GENERAL DISCUSSION - PATHOPHYSIOLOGICAL CONSIDERATIONS

LIVER ENZYMES IN RELATION TO METABOLIC SYNDROME, TYPE 2 DIABETES MELLITUS AND CARDIOVASCULAR RISK

This chapter is based on: Schindhelm RK, Diamant M, Heine RJ. Non-alcoholic fatty

liver disease and cardiovascular disease risk. Current Diabetes Reports 2007.

With permission from Current Science Inc., Philadelphia, USA.

CHAPTER 10

178

INTRODUCTION

Non-alcoholic fatty liver disease (NAFLD) is characterized by a wide spectrum of

liver abnormalities ranging from liver steatosis to the more severe non-alcoholic

steatohepatitis (NASH), resembling alcohol-induced liver disease. By definition

NAFLD develops in subjects who are not heavy alcohol consumers and who have

negative tests for viral and autoimmune liver diseases. It usually has a benign

clinical course, but it may progress to NASH, fibrosis, cirrhosis and rarely, to

hepatocellular carcinoma [1,2]. In recent years, NAFLD has gained appreciation

because of its relationship with insulin resistance, metabolic syndrome and type 2

diabetes mellitus (DM2). Currently, NAFLD is considered by some authors to be the

hepatic component of the metabolic syndrome [3,4], and evidence is accumulating

that patients with NAFLD are at increased risk of developing cardiovascular disease

(CVD).

NAFLD IN RELATION TO THE METABOLIC SYNDROME AND DM2

The putative role of the liver in the pathogenesis of DM2 has gained much interest

and several cross-sectional studies have demonstrated that NAFLD is related to

features of the metabolic syndrome and DM2 [5-7]. In the analysis of the third

National Health and Nutrition Examination Survey (NHANES III) up to 31% of the

elevated alanine aminotransferase (ALT) activity could be explained by high alcohol

consumption, hepatitis B or C infection and/or high transferrin saturation, whereas

in the remaining 69%, the elevated ALT activity was significantly associated with

higher body mass index (BMI), waist circumference, triglycerides, fasting insulin

and lower high density lipoprotein (HDL) cholesterol [5]. Several studies have

addressed the prospective relation of ALT and the metabolic syndrome and DM2

[8,9]. In the Hoorn Study, a population-based cohort study among elderly Caucasian

men and women, we found that ALT was associated with the development of the

metabolic syndrome after 6 years of follow-up (Chapter 9). In patients with DM2,

elevated serum ALT enzyme activity is more frequently observed than in the

general population [10,11]. In addition, some [12-16], but not all studies [8,17-19]

(including the study presented in Chapter 9), have demonstrated independent and

significant associations of ALT with future DM2. The observed association between

ALT and incident DM2 in the mentioned studies may be explained by the fact that

they were performed in high-risk populations that may not be representative for

the general population. Overall, these studies show that patients with NAFLD are at

increased risk for developing DM2 and the metabolic syndrome, suggesting that

the increased CVD risk may be mediated via components of the metabolic

GENERAL DISCUSSION

179

syndrome and DM2. Since the pathophysiology linking NAFLD with either the

metabolic syndrome and/or diabetes has not been clarified, it is difficult to make

the distinction between confounding and mediating variables in the

epidemiological analyses.

NAFLD AND INCREASED CVD RISK

Several cross-sectional studies have demonstrated an increase in carotid artery

intima-media thickness (cIMT) in patients with NAFLD [34-36]. However, in these

studies diagnosis of NAFLD was based on evaluation of liver enzymes or on

ultrasound evaluation but not confirmed by liver biopsy, which is regarded as the

“gold standard” in the diagnosis of NAFLD [20]. In a recent study, Targher and

colleagues have demonstrated that patients with biopsy proven NAFLD have a

significantly higher cIMT compared to age-, sex- and BMI matched healthy controls.

Moreover, they demonstrated that the histologically assessed NAFLD predicted

cIMT, independent of classical risk factors, including insulin resistance and

components of the metabolic syndrome [21]. In well-controlled DM2 patients, we

have shown that slightly elevated ALT, as a surrogate marker of NAFLD, is

associated with a decreased brachial artery flow-mediated vasodilatation and an

impaired whole-body insulin sensitivity (Chapter 7). In line with the observation in

patients with DM2, it was shown that non-diabetic patients with NAFLD have a

decreased brachial artery flow-mediated vasodilatation compared to matched

healthy controls, independent of other risk factors, including insulin resistance and

components of the metabolic syndrome [22]. Ioanou and colleagues studied the

association of ALT and the calculated 10-year risk of coronary heart disease as

estimated with the Framingham risk score (FRS). Subjects with elevated ALT values

had a significantly higher FRS compared to those with normal ALT values,

indicating a higher CVD risk in patients with NAFLD [23]. Only a limited number of

studies have addressed the prospective relation of ALT with CVD and mortality. In

the Hoorn Study, a population-based cohort of Caucasian men and women aged 50

to 75 years, we studied the association of ALT at baseline with all-cause mortality,

and incident cardiovascular and coronary heart disease events (CVD and CHD,

respectively) and found a significant association of ALT with incident CHD after

adjustment for components of the metabolic syndrome and traditional CVD risk

factors (Chapter 8). We found no independent associations of ALT with CVD events

and all-cause mortality. The latter result was in line with a previous study by Arndt

and colleagues who studied the association of ALT with all-cause mortality in 8,043

male construction workers and found no significant association [24]. In contrast, a

CHAPTER 10

180

recent study by Nakamura and co-workers found a positive association of ALT with

all-cause mortality in Japanese men and women, but only for those with a body

mass index below the median (22.7 kg/m2) [25].

MECHANISMS LINKING NAFLD TO INCREASED CVD RISK

The increased CVD risk associated with NAFLD might be explained by the close

relation of NAFLD with components of the metabolic syndrome and DM2, although

recent studies support the notion that NAFLD in itself might contribute to the

increased CVD risk [21,22]. However, the mechanisms for this putative relationship

are not clear. Several of the following, probably highly interrelated, factors

contribute to the enhanced risk of DM2 and metabolic syndrome in persons with

NAFLD.

Oxidative stress and inflammation

Hepatic steatosis, promoted by an insulin resistance mediated elevated flux of free

fatty acids (FFA), may worsen to further liver injury as a consequence of oxidative

stress [26] (Figure 10.1). However, the underlying factors that promote disease

progression to cirrhosis are not understood. Enhanced FFA oxidation increases the

formation of oxygen radicals, in turn leading to lipid peroxidation, mitochondrial

dysfunction and cell damage with subsequent release of cytokines, including tumor

necrosis factor-α (TNF-α), interleukin-6 (IL-6) and C-reactive protein (CRP). Indeed,

Kerner and colleagues found an association of ALT with CRP [27]. In contrast,

Haukeland and coworkers studied the role of systemic inflammation in individuals

with NAFLD and NASH compared to healthy controls and demonstrated that IL-6,

but not CRP, after adjustment for BMI, age and sex, was elevated in NAFLD

compared to controls and TNF-α was higher in NASH as compared to NAFLD [28].

Of interest, a study by Targher and coworkers demonstrated that CRP was elevated

in individuals with NAFLD, but this association was largely explained by the amount

of visceral fat [29].

Adiponectin, leptin and resistin

Decreased adiponectin levels, may represent another mechanism linking NAFLD to

CVD. Patients with NAFLD have lower levels of adiponectin compared to healthy

controls, independent of components of the metabolic syndrome [28,30,31]. This

observation may be relevant, as some studies have shown that lower levels of

GENERAL DISCUSSION

181

adiponectin are associated with CVD [32]. Leptin is a cytokine hormone mainly

produced by adipocytes, that regulates food intake and fat metabolism through

actions on the central nervous system. It has been suggested that leptin may act as

one of the proinflammatory regulators in the progression of NAFLD to NASH by up-

regulation of transforming growth factor β [33]. However, a recent study found no

association of leptin with liver disease severity [34]. Resistin is a protein expressed

in adipose tissue and related to insulin resistance in mice [35]. Recently, a

relationship between resistin and NAFLD and NASH was demonstrated in humans,

with higher plasma resistin levels in individuals with NASH [34].

insulin resistance

↑↑↑↑ lipolysis

obesity

↑↑↑↑ portal FFA↑↑↑↑ FFA

mitochondrial

dysfunction

↑↑↑↑ TGVLDL

assembly/secretion

oxidative

stress

impaired

hepatic

Insulin

signaling

↑↑↑↑ hepatic glucose

↑↑↑↑ de novo

lipogenesis

TG-rich VLDL

glucose

hyperglycemia

[↑↑↑↑ALT]

ALT

hepatocyte

damage

cytokines

adipokines

inflammatory

cytokines

(e.g. CRP)

insulin resistance

↑↑↑↑ lipolysis

obesity

↑↑↑↑ portal FFA↑↑↑↑ FFA

mitochondrial

dysfunction

↑↑↑↑ TGVLDL

assembly/secretion

oxidative

stress

impaired

hepatic

Insulin

signaling

↑↑↑↑ hepatic glucose

↑↑↑↑ de novo

lipogenesis

TG-rich VLDL

glucose

hyperglycemia

[↑↑↑↑ALT]

ALT

hepatocyte

damage

cytokines

adipokines

inflammatory

cytokines

(e.g. CRP)

Figure 10.1. Overview of mechanisms involved in the development of non-alcoholic fatty liver disease (NAFLD)

and the contribution of NAFLD to cardiometabolic risk. NAFLD is characterized by an increased uptake of free

fatty acids (FFA) by the liver, an increased de novo lipogenesis, impaired β-oxidation caused by mitochondrial

dysfunction and an insufficient assembly and secretion of very-low-density lipoproteins (VLDL). The impaired

mitochondrial β-oxidation enhances formation of reactive oxygen species causing hepatocyte damage and up-

regulation of pro-inflammatory cytokines. Oversupply of FFA may lead to excess TG storage and hepatic insulin

resistance due to ser-phosphorylation of components of the insulin-signaling cascade. Hepatic insulin

resistance and increased FFA supply will increase gluconeogenesis, collectively leading to an increased hepatic

glucose output. Alanine aminotransferase (ALT) is released upon hepatocyte damage and possibly up-regulated

by the hepatic insulin resistant state.

CHAPTER 10

182

Postprandial dysmetabolism

Studies comparing the postprandial response of triglycerides and FFAs to a fat-rich

meal in non-diabetic subjects with biopsy proven NASH to controls, showed that

patients with NASH had significantly higher postprandial triglyceride levels than

healthy controls [36,37]. Another study demonstrated that patients with ultrasound

diagnosed NAFLD with an abnormal ALT and or AST had higher glucose levels after

a 75-g oral glucose tolerance test than those with normal ALT and AST [38]. Toledo

and colleagues observed that in obese patients with DM2, increased hepatic

steatosis, as quantified by computed tomography scanning, was positively

correlated with serum triglycerides and inversely with HDL-cholesterol. No

difference in LDL-cholesterol and apolipoprotein B100 was observed relative to the

degree of hepatic steatosis. However, the LDL particle size was smaller in the group

with severe hepatic steatosis. These observations indicate that hepatic steatosis

may contribute to the increased CVD risk in these patients by increasing

triglyceride enrichment of VLDL particles, lowering HDL-cholesterol and by

increasing the number of small, dense LDL-particles. The relationship of hepatic

steatosis with serum triglycerides was stronger in subjects with a minor degree of

hepatic steatosis and weaker in those with a more severe degree of hepatic

steatosis, which may, as the authors suggested, indicate that the incorporation of

triglycerides into VLDL has a limited capacity [39]. In the Hoorn Prandial study, we

found no significant association between ALT and postprandial glucose and

triglycerides (Chapter 2), in contracts to a recent study form our group, in which

we found a significant association between liver fat (quantified by proton-magnetic-

resonance-spectroscopy) and postprandial triglycerides (Tushuizen et al,

unpublished data). These observations suggest that the latter method may be a

better indicator of liver fat in the associations of liver fat with postprandial

dysmetabolic changes than ALT.

CONCLUSIONS

Evidence is accumulating that NAFLD is associated with cardiovascular risk factors

and markers of sub-clinical atherosclerosis, but also to overt CVD events. This

increased risk for CVD necessitates the evaluation and treatment of these patients.

Long-term outcome studies need to establish the benefit of treatment of NAFLD on

the reduction and prevention of diabetes and CVD risk.

GENERAL DISCUSSION

183

REFERENCES


2. Day CP. Genes or environment to determine alcoholic liver disease and non-

alcoholic fatty liver disease. Liver Int 2006;26:1021-28.



4. Marchesini G, Brizi M, Bianchi G et al. Nonalcoholic fatty liver disease: a feature of




6. Ruhl CE., Everhart JE. Determinants of the association of overweight with elevated

serum alanine aminotransferase activity in the United States. Gastroenterology

2003;124:71-9.

7. Ioannou GN, Weiss NS, Boyko EJ, Kahn SE, Lee SP. Contribution of metabolic factors

to alanine aminotransferase activity in persons with other causes of liver disease.



metabolic syndrome and type 2 diabetes in middle-aged Japanese men. Diabetes

Care 2004;27:1427-32.

9. Hanley AJ, Williams K, Festa A, Wagenknecht LE, D'Agostino RB Jr, Haffner SM. Liver

markers and development of the metabolic syndrome: the insulin resistance

atherosclerosis study. Diabetes 2005;54:3140-7.




aminotransferase activity among Mexican Americans. Am J Epidemiol

1997;146:565-71.

12. Ohlson LO, Svardsudd K, Welin L, et al. Fasting blood glucose and risk of coronary

heart disease, stroke, and all-cause mortality: a 17-year follow-up study of men

born in 1913. Diabet Med 1986;3:33-7.







Diabetes 2004;53:2855-60.


type 2 diabetes: the insulin resistance atherosclerosis study. Diabetes

2004;53:2623-32.

CHAPTER 10

184

16. Wannamethee SG, Shaper AG, Lennon L, Whincup PH. Hepatic enzymes, the

metabolic syndrome, and the risk of type 2 diabetes in older men. Diabetes Care

2005;28:2913-8.

17. Nannipieri M, Gonzales C, Baldi S, et al. Liver enzymes, the metabolic syndrome,

and incident diabetes: the Mexico City diabetes study. Diabetes Care

2005;28:1757-62.

18. Schindhelm RK, Dekker JM, Nijpels G, Heine RJ, Diamant M. No independent

association of alanine aminotransferase with risk of future type 2 diabetes in the

Hoorn study. Diabetes Care 2005;28:2812.

19. Andre P, Balkau B, Born C, et al. Hepatic markers and development of type 2

diabetes in middle aged men and women: a three-year follow-up study. Diabetes

Metab 2005;31:542-50.

20. Hubscher SG. Histological assessment of non-alcoholic fatty liver disease.

Histopathology 2006;49:450-65.

21. Targher G, Bertolini L, Padovani R, Zenari L, Zoppini G, Falezza G. Relation of

nonalcoholic hepatic steatosis to early carotid atherosclerosis in healthy men: role

of visceral fat accumulation. Diabetes Care 2004;27:2498-500.

22. Villanova N, Moscatiello S, Ramilli S, et al. Endothelial dysfunction and

cardiovascular risk profile in nonalcoholic fatty liver disease. Hepatology

2005;42:473-80.

23. Ioannou GN, Weiss NS, Boyko EJ, Mozaffarian D, Lee SP. Elevated serum alanine

aminotransferase activity and calculated risk of coronary heart disease in the United

States. Hepatology 2006;43:1145-51.

24. Arndt V, Brenner H, Rothenbacher D, Zschenderlein B, Fraisse E, Fliedner TM.

Elevated liver enzyme activity in construction workers: prevalence and impact on

early retirement and all-cause mortality. Int Arch Occup Environ Health

1998;71:405-12.

25. Nakamura K, Okamura T, Kanda H, Hayakawa T, Okayama A., Ueshima H. The value

of combining serum alanine aminotransferase levels and body mass index to

predict mortality and medical costs: a 10-year follow-up study of national health

insurance in Shiga, Japan. J Epidemiol 2006;16:15-20.

26. Videla LA, Rodrigo R, Araya J, Poniachik J. Insulin resistance and oxidative stress

interdependency in non-alcoholic fatty liver disease. Trends Mol Med 2006;12:555-

8.




28. Haukeland JW, Damas JK, Konopski Z, et al. Systemic inflammation in nonalcoholic

fatty liver disease is characterized by elevated levels of CCL2. J Hepatol

2006;44:1167-74.

GENERAL DISCUSSION

185

29. Targher G, Bertolini L, Scala L, Zoppini G, Zenari L, Falezza G. Non-alcoholic hepatic

steatosis and its relation to increased plasma biomarkers of inflammation and

endothelial dysfunction in non-diabetic men. Role of visceral adipose tissue. Diabet

Med 2005;22:1354-8.

30. Pagano C, Soardo G, Esposito W, et al. Plasma adiponectin is decreased in

nonalcoholic fatty liver disease. Eur J Endocrinol 2005;152:113-8.

31. Targher G, Bertolini L, Rodella S, et al. Associations between plasma adiponectin

concentrations and liver histology in patients with nonalcoholic fatty liver disease.

Clin Endocrinol (Oxf) 2006;64:679-683.

32. Pischon T, Girman CJ, Hotamisligil GS, Rifai N, Hu FB, Rimm EB. Plasma adiponectin

levels and risk of myocardial infarction in men. JAMA 2004;291:1730-7.

33. Ikejima K, Okumura K, Lang T, et al. The role of leptin in progression of non-

alcoholic fatty liver disease. Hepatology Research 2005;33:151-4.

34. Pagano C, Soardo G, Pilon C, et al. Increased serum resistin in nonalcoholic fatty

liver disease is related to liver disease severity and not to insulin resistance. J Clin

Endocrinol Metab 2006;91:1081-6.

35. Muse ED, Obici S, Bhanot S, et al. Role of resistin in diet-induced hepatic insulin

resistance. J Clin Invest 2004;114:232-9.

36. Cassader M, Gambino R, Musso G, et al. Postprandial triglyceride-rich lipoprotein

metabolism and insulin sensitivity in nonalcoholic steatohepatitis patients. Lipids

2001;36:1117-24.

37. Musso G, Gambino R, De Michieli F, et al. Dietary habits and their relations to

insulin resistance and postprandial lipemia in nonalcoholic steatohepatitis.

Hepatology 2003;37:909-16.

38. Su CC, Wang K, Hsia TL, Chen CS, Tung TH. Association of nonalcoholic fatty liver

disease with abnormal aminotransferase and postprandial hyperglycemia. J Clin

Gastroenterol 2006;40:551-4.

39. Toledo FG, Sniderman AD, Kelley DE. Influence of hepatic steatosis (fatty liver) on

severity and composition of dyslipidemia in type 2 diabetes. Diabetes Care

2006;29:1845-50.

CHAPTER 10

186

187

Appendix

SUMMARY & SAMENVATTING

188

SUMMARY

Chapter 1 starts with a brief outline about diabetes and the metabolic syndrome in

relation to cardiovascular disease (CVD) risk in patients with type 2 diabetes

mellitus (DM2) and introduces some new putative CVD risk determinants that may

explain the excess CVD risk in patients with DM2.

In Chapter 2, we describe the determinants of postprandial triglyceride and

glucose concentrations following two consecutive (breakfast and lunch) fat-rich and

carbohydrate-rich meals in post-menopausal women with normal glucose

metabolism (NGM) and in women with DM2 who were part of the Hoorn Prandial

Study. The postprandial increment in triglycerides did not differ between women

with NGM and DM2, whereas fasting triglycerides and the total triglyceride

response was higher in DM2. Fasting triglyceride concentration, HbA1c, total

cholesterol, and, inversely, high density lipoprotein (HDL) cholesterol were

independently associated with the postprandial increment in triglycerides in women

with NGM. In women with DM2, the fasting triglyceride concentration was the only

determinant of the postprandial increment in triglycerides. In women with NGM,

age and fasting triglycerides were independently associated with postprandial

increment in glucose. In women with DM2, HbA1c was independently associated

with the postprandial increment in glucose, and in women on statin therapy

physical activity was an additional determinant. We concluded that the postprandial

increment in triglycerides did not differ between post-menopausal women with

NGM and well-controlled DM2. Furthermore, fasting triglyceride levels were a

strong predictor of the postprandial increment in triglycerides, whereas the

postprandial increment in glucose was especially determined by variables other

than fasting glucose.

The study in Chapter 3 compared the associations of postprandial glucose and

postprandial triglycerides with carotid intima media thickness (cIMT) after two

consecutive fat-rich or carbohydrate rich meals in women with NGM and DM2. In

women with NGM, an increase of 1.0 mmol/l in plasma glucose following the fat-

rich meals and an increase of 1.8 mmol/l in plasma glucose concentration

following the carbohydrate meals were associated with a 50 µm larger cIMT. No

association between postprandial glucose and cIMT was found in women with DM2.

Postprandial triglycerides were not associated with cIMT. The association between

postprandial glucose and cIMT in women with NGM suggests that postprandial

glucose in the normal range is a marker or a risk factor for atherosclerosis.


189

In the next chapter (Chapter 4), we addressed possible mechanisms that link

postprandial dysmetabolism to CVD risk. The influence of two consecutive meals

(breakfast and lunch) with different composition (both fat-rich or carbohydrate-rich)

on acute changes in various markers of glycoxidative and lipoxidative stress,

including oxidized low-density-lipoprotein (oxLDL), Nε-(carboxyethyl)-lysine (CEL),

Nε-(carboxymethyl)-lysine (CML) and 3-deoxyglucosone (3DG) were studied in a

sub-sample of the Hoorn Prandial Study (27 with NGM and 26 with DM2). We found

significant elevations of oxLDL, 3DG and CML during the postprandial period in

post-menopausal women with DM2 and NGM. The fasting values of oxLDL, 3DG

and CML in women with DM2 were of similar magnitude as the postprandial values

in women with NGM. The elevations of oxLDL were closely correlated with changes

in postprandial triglycerides, whereas 3DG alterations were correlated with changes

in postprandial glucose following the fat-rich meals. CML and CEL correlated to with

neither postprandial glucose nor triglyceride changes. These results suggest that

exaggerated glycoxidative and lipoxidative stress, due to postprandial

dysmetabolism, may contribute to the increased CVD risk in DM2.

In Chapter 5 we studied the 8-hour time course of myeloperoxidase (MPO)

following two consecutive (breakfast and lunch) fat-rich and carbohydrate rich-

meals in postmenopausal women with NGM and in women with DM2. MPO is

abundantly expressed in leukocytes, and released upon activation in the

extracellular space. MPO has been associated with CVD and endothelial

dysfunction. Postprandial leukocyte recruitment and activation with subsequent

MPO release may contribute to atherosclerosis and CVD. We hypothesized that

plasma MPO would increase in the postprandial state, due to leukocyte recruitment

and/or activation, especially in individuals with DM2. Baseline MPO was not

significantly different between NGM and DM2. Baseline MPO was weakly associated

with leukocytes and inversely associated with HDL-cholesterol. In the postprandial

phase, in contrast to our hypothesized increase, MPO decreased in NGM (both meal

types) and in DM2 (fat-rich meals). Based on these results we concluded that our

findings provide no support to our initial hypothesis that meal induced release of

MPO might be a mechanism that contributes to CVD risk in DM2.

Chapter 6 reviews the applicability of alanine aminotransferase (ALT) as a marker

of non-alcoholic fatty liver disease (NAFLD) and provides an overview of the

epidemiological studies that assessed the associations between ALT and the

metabolic syndrome, DM2 and CVD. We concluded that in epidemiological studies,

ALT is an adequate marker of NAFLD, provided that alcohol-intake as a


190

confounding or effect-modifying variable is considered, and provided that other

more rare causes of ALT elevation are excluded.

In a cross-sectional study in 64 normotriglyceridemic patients with DM2, we

studied the relation of ALT with endothelial function, measured as flow-mediated

vasodilatation (FMD), and with insulin sensitivity (M/I-value) (Chapter 7). We found

that ALT was significantly, albeit weakly, associated with FMD and M/I-value,

independent of obesity. Moreover, the relation of ALT with FMD was independent of

insulin sensitivity. These findings suggest that in DM2, elevated ALT is associated

with a more unfavorable CVD risk profile.

In the study described in Chapter 8 we used data from the Hoorn Study to assess

the prospective relation of ALT with all-cause mortality, CVD and coronary heart

disease (CHD) events in a population-based cohort of elderly Caucasian men and

women. We demonstrated that individuals with ALT levels in the upper tertile, had a

significantly higher risk for CHD events that those in the lower tertile, after

adjustment for components of the metabolic syndrome and traditional risk factors.

No significant associations were found between ALT and all-cause mortality and

CVD events, after applying the same adjustment. This study shows that individuals

with even slightly elevated ALT levels are at an increased risk for CHD events. The

pathogenic mechanisms underlying this association are not incompletely

understood and require further study.

In Chapter 9, we studied the prospective relation of ALT with incident metabolic

syndrome and DM2. We found that higher ALT levels were related to conversion to

the metabolic syndrome, after adjustment for confounding factors including

alcohol-intake, age and the individual components of the metabolic syndrome.

However, in contrast to previous reports, we could not demonstrate an

independent prospective relation between ALT and the risk of DM2. These results

suggest that the increased CVD risk might be mediated by metabolic derangements

that differ from those that result in hyperglycemia (Chapter 9).

The last chapter (Chapter 10) discusses the methodological and pathophysiological

aspects of the studies presented in this thesis.


191

SAMENVATTING VOOR NIET INGEWIJDEN

Het aantal gevallen van “ouderdomssuikerziekte” (type 2 diabetes mellitus) neemt

in de Westerse wereld snel toe als gevolg van veranderingen in leefstijl (verhoogde

consumptie van vet- en energierijke voeding en verminderde lichaamsbeweging),

en als gevolg van de vergrijzing van de bevolking. Mensen met type 2 diabetes

mellitus hebben een verhoogd risico op het krijgen van hart- en vaatziekten. Dit

verhoogde risico kan voor een deel verklaard worden door reeds bekende

risicofactoren voor hart- en vaatziekten, zoals hoge bloeddruk (hypertensie), roken,

afwijkende vetspiegels in het bloed (een verhoogd LDL-cholesterol (het “slechte”

cholesterol), laag HDL cholesterol (goede cholesterol) en verhoogde hoeveelheden

triglyceriden), alsook hoge glucose (suiker) concentraties in het bloed. Met name

bij vrouwen na de menopauze (overgang) is het risico op hart- en vaatziekten hoger

dan bij vrouwen vóór de menopauze. Na de overgang stijgt de hoeveelheid

triglyceriden en daalt de hoeveelheid HDL-cholesterol in het bloed. Dit heeft ons

aangespoord de effecten van maaltijden te bestuderen bij vrouwen na menopauze.

Het is al langer bekend dat glucose en triglyceriden, en verbindingen die hieruit

gevormd kunnen worden, schadelijk kunnen zijn voor de bloedvaten en daardoor

kunnen bijdragen aan het ontwikkelen van hart- en vaatziekten. Ruim een kwart

eeuw geleden is geopperd dat verstoringen in de vet- en glucosestofwisseling die

optreden na een maaltijd mogelijk kunnen bijdragen aan het verhoogde risico op

hart- en vaatziekten. Na het eten van een maaltijd stijgen de triglyceriden- en

glucoseconcentraties in het bloed. De triglyceriden en de glucose worden gebruikt

als bouwstoffen of als brandstoffen voor het lichaam. De mate waarin deze stoffen

worden verwerkt en opgeslagen verschilt tussen gezonde mensen en mensen met

type diabetes mellitus. De lever speelt hierbij een belangrijke rol. Een verhoogde

hoeveelheid vet in de lever (steatose), een aandoening die vaker voorkomt bij

mensen met diabetes en overgewicht, kan deze processen verstoren, alsook het

risico op hart- en vaatziekten verhogen. Het hormoon insuline speelt een

belangrijke rol bij het verwerken van de triglyceriden en de glucose. Bij mensen

met type 2 diabetes mellitus werkt het insuline minder goed waardoor de

triglyceriden en de glucose trager worden opgenomen en verwerkt. De vraag is

welke stof schadelijker is voor de vaatwand: de glucose of de triglyceriden?

De onderzoeken die in dit proefschrift beschreven zijn, proberen antwoord te

geven op de volgende vragen: 1a) hoe is het beloop van glucose en triglyceriden in

het bloed na twee vetrijke en twee koolhydraatrijke maaltijden, 1b) wat is de

relatieve bijdrage van glucose en triglyceriden in relatie tot hart- en vaatziekten en


192

1c) welke mechanismen kunnen het verhoogde risico op hart- en vaatzieken

verklaren bij patiënten met type 2 diabetes mellitus; 2a) hebben mensen met

leververvetting een verhoogde kans op het risico van hart- en vaatzieken en 2b) wat

zijn de mogelijke mechanismen waardoor dit verklaard kan worden?

In Hoofdstuk 1 geven we een korte beschrijving van type 2 diabetes mellitus en

het metabool syndroom (cluster van risicofactoren samenhangend met

overgewicht). Daarnaast geven we een kort overzicht van bekende en nieuwe

determinanten voor hart- en vaatziekten bij mensen met type 2 diabetes mellitus.

Tenslotte beschrijft dit hoofdstuk de vraagstellingen en de opbouw van dit

proefschrift.

Hoofdstuk 2 beschrijft het beloop van glucose en triglyceriden in het bloed na

twee vetrijke dan wel koolhydraatrijke maaltijden in post-menopauzale vrouwen

met en zonder type 2 diabetes mellitus. Ongeveer de helft van de vrouwen met

type 2 diabetes mellitus werd behandeld met cholesterolverlagende medicatie

(statines) en gezien het mogelijke effect op de vetstofwisseling werden de

resultaten voor deze vrouwen afzonderlijk gepresenteerd. Bij vrouwen met diabetes

(met en zonder statines) waren de hoeveelheid triglyceriden en glucose hoger dan

bij de gezonde vrouwen. We vonden echter geen verschil in de toename van

triglyceriden na de vetrijke maaltijden tussen de vrouwen met diabetes en de

gezonde vrouwen. Bij de gezonde vrouwen werd een verband gezien tussen de

stijging van de triglyceriden concentratie na de vetrijke maaltijden en de nuchtere

triglyceriden waarde, het totaal cholesterol, het HbA1c (afspiegeling van de

gemiddelde hoeveelheid glucose in het bloed gedurende 6-8 weken) en het HDL-

cholesterol. Bij vrouwen met diabetes was de nuchtere triglyceride waarde de

belangrijkste bepalende factor voor de triglyceride waarden na beide vetrijke

maaltijden. De glucosewaarde na de koolhydraatrijke maaltijden werd bij de

gezonde vrouwen voornamelijk bepaald door leeftijd en de nuchtere

triglyceridenwaarde en in de vrouwen met diabetes door HbA1c.

Het onderzoek naar het mogelijke verband tussen de bijdrage van de nuchtere en

post-prandiale (na de maaltijd) glucose en triglyceridenconcentraties en

manifestaties van hart- en vaatziekten is beschreven in Hoofdstuk 3. De dikte van

de vaatwand (intima-media dikte) van de halsslagader gemeten met een

echotechniek werd gebruikt als maat voor het risico op hart- en vaatziekten. We

vonden dat postprandiale glucosewaarden gerelateerd was aan de intima-media

dikte bij de gezonde vrouwen en niet bij de vouwen met type 2 diabetes mellitus.


193

De postprandiale triglyceridenconcentraties waren niet gerelateerd aan de intima-

media dikte. We concludeerden dat de glucosespiegel na de maaltijd een indicator

kan zijn voor hart- en vaatziekten bij vrouwen zonder type 2 diabetes mellitus.

In Hoofdstuk 4 bestudeerden we het effect van twee vetrijke en van twee

koolhydraatrijke maaltijden op de vorming van geoxideerde LDL-cholesterol

deeltjes (oxLDL), de reactieve suikerverbinding 3-deoxyglucoson (3DG) en twee

verschillende versuikerde eiwitverbindingen (de zogenaamde ‘advanced glycation

end-products’ CML en CEL). Van deze verschillende verbindingen wordt

verondersteld dat ze mogelijk een relatie hebben met hart- en vaatziekten. Over de

vorming van deze verbindingen na de maaltijd is weinig bekend. We vonden dat de

nuchtere waarden van oxLDL, 3DG en CML hoger waren bij de vrouwen met type 2

diabetes mellitus dan bij de gezonde vrouwen. De nuchter waarden van deze

verbindingen bij vrouwen met type 2 diabetes waren vergelijkbaar met de

postprandiale waarden bij de gezonde vrouwen. Het oxLDL correleerde met de

hoeveelheid postprandiale triglyceriden na de vetrijke maaltijden en het 3DG

correleerde met hoeveel postprandiale glucose na de vetrijke maaltijd. We

concludeerden dat oxLDL, 3DG en CML mogelijk kunnen bijdragen aan het

voorhoogde risico op hart- en vaatziekten.

In Hoofdstuk 5 besschrijven we het effect van twee koolhydraatrijke en twee

vetrijke maaltijden op de vorming van myeloperoxidase. Dit is een enzym dat in

witte bloed cellen (leukocyten) wordt gemaakt en dat betrokken is bij de afweer van

infecties. Een aantal studies hebben een relatie laten zien met hart- en vaatzieken.

We onderzochten of dit enzym toe neemt na de maaltijden en in het bijzonder bij

de vrouwen met type 2 diabetes mellitus. Tot onze verbazing zagen we, ondanks

de stijging van de leukocyten, geen toename, maar juist een afname van het enzym

na beide maaltijdtypen. Dit gebeurde zowel bij de gezonde vrouwen als bij de

vrouwen met type 2 diabetes mellitus. We concludeerden dat myeloperoxidase na

de maaltijd geen belangrijk factor is in relatie tot hart- en vaatziekten.

Hoofdstuk 6 introduceert de niet-alcoholische leververvetting, een vorm van

leververvetting die optreedt bij mensen die niet of matig alcohol nuttigen. Deze

vorm wordt vaak gezien bij mensen met overgewicht en type 2 diabetes mellitus. In

de onderzoeken die in dit proefschrift beschreven zijn, wordt het leverenzym

alanine aminotransferase (ALAT) gebruikt als indicator voor leververvetting. Dit

hoofdstuk beschrijft de toepasbaarheid van ALAT in relatie tot leververvetting.


194

In Hoofdstuk 7 vonden we dat bij goed ingestelde patiënten met type 2 diabetes

mellitus, een hoge ALAT waarde in het bloed, als marker voor leververvetting,

samenhangt met een verminderde endotheel-afhankelijke vaatverwijding (geeft een

indruk over de werking van het endotheel, de “binnenbekleding” van de bloedvaten)

en een verminderde insulinegevoeligheid. Deze bevinding suggereert dat mensen

met leververvetting een verhoogde kans hebben op hart- en vaatziekten. Alanine

aminotransferase, endotheel-afhankelijke vaatverwijding en de insuline

gevoeligheid zijn echter op één moment gemeten, waardoor er geen uitspraak

gedaan kan worden over oorzaak en gevolg.

De vraag of een verhoogde ALAT waarde gepaard gaat met het optreden van hart-

en vaatzieken wordt beantwoord in Hoofdstuk 8. In de Hoorn Studie (een

bevolkingsonderzoek van mannen en vrouwen tussen 50 en 75 jaar) vonden wij dat

een hoge ALAT waarde, als marker voor leververvetting, samenhangt met het

optreden van kransslagaderaandoeningen (coronaire hartziekten) gedurende een

vervolgperiode van 11 jaar. Deze relatie kon niet verklaard worden door reeds

bekende risicofactoren. Onze conclusie was dan ook dat mensen met

leververvetting een verhoogd risico hebben op coronaire hartziekten.

Hoofdstuk 9 beschrijft de rol van ALAT als voorspeller (na 6 jaar) voor het

optreden van type 2 diabetes mellitus en het metabool syndroom in de Hoorn

Studie. We vonden dat ALAT type 2 diabetes voorspelt, maar deze relatie kon deels

verklaard worden door reeds bekende risicofactoren voor diabetes mellitus, zoals

een licht verhoogde hoeveelheid glucose in het bloed en het hebben van

overgewicht. ALAT voorspelde wel het optreden van het metabool syndroom.

In het laatste hoofdstuk (Hoofdstuk 10) worden de bevindingen van dit

proefschrift bediscussieerd, de relevantie voor de klinische praktijk beschreven en

worden suggesties gedaan voor verder onderzoek op dit gebied.


195

DANKWOORD

196

DANKWOORD

Velen hebben bijgedragen aan de totstandkoming van dit proefschrift en graag wil ik hen

danken voor hun bijdrage.

De deelnemers van de Hoorn VeT studie, de Hoorn studie en de “triglyceriden-thuismeet”-

studie, wil ik bedanken voor hun deelname: zonder uw medewerking geen

wetenschappelijk onderzoek!

Mijn promotor, prof. dr. R. J. Heine, beste Rob, dank voor de mogelijkheid en de ruimte om

dit onderzoek te doen, en je aanstekelijk enthousiasme over alles wat met

diabetes(onderzoek) te maken heeft.

Mijn copromotor, dr. M. Diamant, beste Michaela, wat een energie en enthousiasme heb jij,

jouw kennis is ongeëvenaard en ik ben zeer vereerd om één van jouw eerste promovendi te

mogen zijn. Dank voor je begeleiding en de nodige pep-talks, ik heb heel veel van je

geleerd!

Mijn copromotor, dr. T. Teerlink, beste Tom, jouw commentaren op mijn manuscripten

lagen meestal al eerste in mijn email-box, dank voor al je ondersteuning en meedenken bij

de Hoorn VeT studie en de (vele) pilotstudies.

Dr. M. Alssema, beste Marjan, samen hebben we deze onderzoeks-“klus” geklaard, dank

voor de geweldige samenwerking en de gezelligheid. Ik vind het een enorme eer dat je

mijn paranimf wilt zijn. Heel veel succes en inspiratie in je verdere carrière!

De Hoorn VeT studie was zeker niet mogelijk geweest zonder de enorme inzet van Jannet

Entius, Tanja Nansink, Lida Ooteman and Marianne Veeken, die zorg droegen voor het

versturen van de brieven, bellen van de deelnemers, invriezen van de enorme hoeveelheid

bloedmonsters,… en de vele gezellige momenten tijdens het “veld-werk” in het Diabetes

Onderzoek Centrum in Hoorn. Jannet, Tanja, Lida en Marianne, heel veel dank voor jullie

inzet! Marianne, super dat je mijn paranimf wilt zijn!

DANKWOORD

197

Jolanda Bosman, dank voor je hulp en meedenken in het opzetten en plannen van de

Hoorn VeT studie binnen het Diabetes Onderzoek Centrum in Hoorn.

De overige begeleiders van de Hoorn VeT studie: prof. dr. ir. J. M. Dekker, beste Jacqueline,

dank voor al je input bij de Hoorn VeT studie, en je hulp bij het analyseren van de Hoorn

Studie data. Prof. dr. G. Nijpels, beste Giel, dank voor je input en je commentaren op mijn

manuscripten. Dr P. J. Kostense, beste Piet, dank voor het meedenken bij de opzet van de

Hoorn VeT studie en je veelvuldige uitleg over de meest (in mijn ogen dan) ingewikkelde

statistische technieken.

Dr. P. G. Scheffer, beste Peter, dank voor al je (lab-technische) input, en je begeleiding bij

het MPO-stuk. Ik ben vereerd dat je plaats neemt in de leescommissie en zeer benieuwd

naar je opmerkingen en vragen tijdens de verdediging.

De overige leden van de leescommissie: prof. dr. J. C. Seidell, prof. dr. Y. M. Smulders,

prof. dr. U. Smith en dr. E. S. G. Stroes. Ik dank u voor het beoordelen van het manuscript

van mijn proefschrift en de tijd die u hiervoor heeft vrijgemaakt.

Prof. dr. L. M. Bouter en prof. dr. C. D. A. Stehouwer, dank voor uw snelle en vaak

kleurrijke commentaren op mijn Hoorn Studie manuscripten. Ik ben er veel van mogen

leren.

De overige co-auteurs, dr. S. J. L. Bakker, dr. R.A. van Dijk, dr. C. G. Schalkwijk en drs. M.E.

Tushuizen, beste Stephan, Rob, Casper en Maarten, dank voor jullie bijdragen!

Danielle van Assema, dank voor je bijdrage aan de 3DG-pilot studie.

De medewerkers van de afdeling Klinische Chemie, in het bijzonder Rob Barto, Sigrid de

Jong, Jeany Huijser, Ina Kamphuis, Astrid Kok, Harry Twaalfhoven, Rick Vermue: dank voor

jullie ondersteuning en/of het uitvoeren van de vele laboratoriumbepalingen.

Alle oud-collega’s van de afdeling Endocrinologie, het EMGO instituut en het Diabetes

Onderzoek Centrum wil ik danken voor de prettige samenwerking en de gezellige

momenten. Annemarie, Carolien, Daniël, Esther, Josina, Hata, Karlijn, Katja, Laura, Luuk,

DANKWOORD

198

Maarten, Marieke, Mathijs, Natalia, Nynke, Renate, Sigridur en Wiebe, heel veel succes met

(het afronden van) jullie (promotie)onderzoeken!

Mijn lieve familie, schoonfamilie en vrienden wil ik bedanken voor hun belangstelling en

alle gezeligheid naast het onderzoek.

Pap en mam, heel veel dank voor jullie onvoorwaardelijk liefde en steun om te worden wie

ik nu ben.

Lieve Maurice, wat zou ik zijn zonder jou. Dank voor al je steun en liefde. Op naar ons

volgende avontuur… Ti amo da morire!

LIST OF PUBLICATIONS

200


1. Schindhelm RK, Hospers HJ. Sex with men before coming-out: relation to sexual activity

and sexual risk-taking behavior. Arch Sex Behav 2004;33:585-91.

2. Schindhelm RK, Diamant M, Bakker SJ, Van Dijk RA, Scheffer PG, Teerlink T, Kostense PJ,

Heine RJ. Liver alanine aminotransferase, insulin resistance and endothelial dysfunction in

normotriglyceridaemic subjects with type 2 diabetes mellitus. Eur J Clin Invest

2005;35:369-74.

3. Schindhelm RK, Dekker JM, Nijpels G, Heine RJ, Diamant M. No independent association

of alanine aminotransferase with risk of future type 2 diabetes in the Hoorn Study (letter).

Diabetes Care 2005;28:2812.

4. Schindhelm RK, Hospers HJ. Mannen die seks hebben met mannen vóór de coming-out:

relatie met seksuele activiteit en seksueel risicogedrag. Tijdschrift voor Seksuologie

2006;30:10-5.

5. Schindhelm RK, Diamant M, Dekker JM, Tushuizen ME, Teerlink T, Heine RJ. Alanine

aminotransferase as a marker of non-alcoholic fatty liver disease in relation to type 2

diabetes mellitus and cardiovascular disease (review). Diabetes Metab Res Rev

2006;22:437-43.

6. Schindhelm RK, Diamant M, Heine RJ. Non-alcoholic fatty liver disease as a determinant of

cardiovascular disease: response to Targher (letter). Atherosclerosis 2007;190:20-21.

7. Schindhelm RK, Dekker JM, Nijpels G, Bouter LM, Stehouwer CD, Heine RJ, Diamant M,

Alanine aminotransferase predicts coronary heart disease events: a 10-year follow-up of

the Hoorn Study. Atherosclerosis 2007;191:391-396.

8. Schindhelm RK, Dekker JM, Nijpels G, Stehouwer CD, Bouter LM, Heine RJ, Diamant M.

Alanine aminotransferase and the 6-year risk of the metabolic syndrome in Caucasian men

and women: The Hoorn Study. Diabet Med 2007;24;430-435.

9. Schindhelm RK, Diamant M, Heine RJ. Nonalcoholic fatty liver disease and cardiovascular

disease risk. Curr Diab Rep (in press).


201

10. Alssema M, Schindhelm RK, Dekker JM, Diamant M, Kostense PJ, Teerlink T, Scheffer PG,

Nijpels, G, Heine RJ. Postprandial glucose, and not triglyceride concentrations are

associated with carotid intima media thickness in women with normal glucose

metabolism: the Hoorn Prandial Study. Atherosclerosis (in press).

11. Schindhelm RK, Alssema M, Scheffer PG, Diamant M, Dekker JM, Bart R, Nijpels G,

Kostense PJ, Heine RJ, Schalkwijk CG, Teerlink T. Fasting and postprandial glycoxidative

and lipoxidative stress are increased in women with type 2 diabetes mellitus. (submitted).

12. Schindhelm RK, Alssema M, Diamant M, Teerlink T, Dekker JM, Kok A, Kostense PJ,

Nijpels G, Heine RJ, Scheffer PG. Comparison of the effect of two consecutive fat-rich and

carbohydrate-rich meals on levels of postprandial plasma myeloperoxidase in women with

and without type 2 diabetes mellitus. (submitted).

13. Alssema M*, Schindhelm RK*, Dekker JM, Diamant M, Nijpels G, Teerlink T, Scheffer PG,

Kostense PJ, Heine RJ. Determinants of postprandial triglyceride and glucose responses

following two consecutive fat-rich or carbohydrate-rich meals in normoglycemic women

and in women with type 2 diabetes: the Hoorn Prandial study. (submitted). (*both authors

contributed equally).

14. Alssema M, Schindhelm RK, Rijkelijkhuizen JM, Kostense PJ, Teerlink T. Nijpels G, Heine

RJ, Dekker JM. Meal-composition affects insulin secretion in women with type 2 diabetes: a

comparison with healthy controls. The Hoorn Prandial Study. (submitted).

203

About the author

Roger Schindhelm was born on February 1st, 1974 in Kerkrade, The Netherlands.

After graduation from secondary school (pre-university education, Sancta Maria

College, Kerkrade) in 1992, he was trained as a biomedical laboratory technician

in clinical chemistry at the university of professional education Limburg and

received his bachelor’s degree in 1996. Subsequently, he studied medicine at

Maastricht University and graduated as a medical doctor (MD) in August 2002.

From October 2002 until December 2006, he was a PhD-student at the Diabetes

Center at the Department of Endocrinology of the VU University Medical Center

in Amsterdam, under supervision of Prof. Robert J. Heine and Drs. Michaela

Diamant and Tom Teerlink. During that period he completed the Postgraduate

Epidemiology Program from the EMGO Institute and received a Master of

Epidemiology degree in 2005.

Over de auteur

Roger Schindhelm is geboren op 1 februari 1974 in het Zuid-Limburge

Kerkrade. Na het afronden van de middelbare school in 1992 (VWO, Sancta

Maria College in Kerkrade), volgde hij de opleiding tot klinisch-chemisch analist

aan de Hogeschool Limburg en behaalde hij zijn HBO-getuigschrift in 1996.

Vervolgens studeerde hij geneeskunde aan de Universiteit Maastricht en slaagde

hij in 2000 voor zijn doctoraalexamen en in 2002 voor zijn artsexamen. Vanaf

oktober 2002 tot en met december 2006 was hij werkzaam als artsonderzoeker

bij het Diabetes Centrum van de afdeling Endocrinologie van het VU medisch

centrum in Amsterdam onder begeleiding van prof. dr. Rob Heine en dr.

Michaela Diamant (afdeling Endocrinologie), en dr. Tom Teerlink (afdeling

Klinische Chemie). Tijdens deze periode volgende hij de postdoctorale opleiding

epidemiologie van het EMGO Instituut welke in 2005 werd afgerond met

registratie tot Epidemioloog A bij de Vereniging voor Epidemiologie.

postprandial dysmetabolism and non-alcoholic fatty liver … · metabolism and the cvd mortality in...

Documents