cardiovascular disease in rheumatoid arthritis367696/s...cardiovascular disease in rheumatoid...

Cardiovascular Disease in Rheumatoid Arthritis

Helen Louise Ruiha Pahau

Diploma (RcompN), PgDip (HSM)

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in 2015

The University of Queensland, Diamantina Institute

i

Abstract

This thesis focuses on cardiovascular (CV) disease (CVD) risk in

rheumatoid arthritis (RA). Inflammation and CV risk factors are increased

in RA patients. This population has a higher carotid intima-media thickness

and more CV events than the general population and kynurenine (KYN)

concentration levels are higher in RA patients compared to healthy controls.

Therefore, in this thesis inflammation and CVD risk factors are investigated

for increased risk of future RA development. RA and type2 diabetes (T2D)

carotid intima-media thickness (CIMT) are compared and measurements of

serum KYN for CVD, cancer and mortality are studied. First I tested the

hypothesis that inflammation, CVD and CV risk factors are associated with

increased risk of future RA development. Data from the Norwegian HUNT

population health survey were obtained to compare groups with RA at

baseline and follow-up (HUNT2 and HUNT3, n = 429) or follow up alone

(HUNT3, n = 786), or without RA at both times (n = 33,567). Results

showed female gender, age, smoking, body mass index (BMI), and history

of previous CVD were associated with self-reported incident RA (previous

CVD: odds ratio 1.52 (95% confidence interval 1.11-2.07). The findings

regarding previous CVD were confirmed in sensitivity analyses excluding

participants with psoriasis (odds ratio (OR) 1.70 (1.232.36)) or restricting

the analysis to cases with hospital diagnosis of RA (OR 1.90 (1.10-3.27)) or

carriers of the shared epitope (OR 1.76 (1.134-2.74)). History of previous

CVD was not associated with increased risk of osteoarthritis (OR 1.04

(0.86-1.27)). I then compared the characteristics of CV risk and progression

of CIMT in 78 RA patients from an outpatient Rheumatology Clinic and

212 patients with T2D who attended a community or hospital diabetes clinic

in Brisbane. We found that the burden of risk varied between the 2 cohorts.

RA patients were older, had a higher proportion of smokers, females and

previous CVD. T2D patients had a higher body mass index (BMI), diastolic

blood pressure, triglycerides, lower high density lipoprotein and a higher

statin use. At baseline, CIMT measurements were similar in the RA and

T2D cohorts. In an adjusted linear regression model, RA was significantly

associated with lower CIMT at follow-up. Despite a shorter follow-up, 91 %

ii

of the T2D cohort had increased CIMT at follow-up compared to 54 % of

the RA cohort. In the RA cohort, disease modifying anti-rheumatic drug

(DMARD) use at baseline was associated with significantly lower CIMT

values at follow-up.

Finally we hypothesized that increased KYN in RA patients may predispose

to CVD, infections and cancer. We measured KYN in 129 RA patients

followed for 10 years for development of CVD, cancer and death. Median

KYN concentrations were significantly higher in RA patients than in

controls, but there was large overlap between groups. There were no

variables in our data set that could explain the differences in KYN

concentrations among the RA patients and KYN concentrations did not

predict development of CVD, new malignancies or death. The number of

pack-years of smoking was the only variable associated with death in

logistic or Cox regression analysis. In summary, CVD risk factors are

associated with increased risk of future RA development. The burden of risk

varies in RA and T2D, and CIMT progression is slower for RA than for

T2D. Although serum KYN concentrations are significantly higher in RA

than in controls, they were not associated with CVD, new malignancies or

death.

iii

Declaration by author

This thesis is composed of my original work, and contains no material

previously published or written by another person except where due

reference has been made in the text. I have clearly stated the contribution by

others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole,

including statistical assistance, survey design, data analysis, significant

technical procedures, professional editorial advice, and any other original

research work used or reported in my thesis. The content of my thesis is the

result of work I have carried out since the commencement of my research

higher degree candidature and does not include a substantial part of work

that has been submitted to qualify for the award of any other degree or

diploma in any university or other tertiary institution. I have clearly stated

which parts of my thesis, if any, have been submitted to qualify for another

award.

I acknowledge that an electronic copy of my thesis must be lodged with the

University Library and, subject to the policy and procedures of The

University of Queensland, the thesis be made available for research and

study in accordance with the Copyright Act 1968 unless a period of

embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides

with the copyright holder(s) of that material. Where appropriate I have

obtained copyright permission from the copyright holder to reproduce

material in this thesis.

Publications during candidature

Publication in peer-reviewed journal

Pahau H, Brown MA, Paul S, Thomas R, Videm V. Cardiovascular disease

is increase prior to onset of rheumatoid arthritis but not osteoarthritis: the

iv

population-based Nord-Trondelag health study (HUNT). Arthritis Res Ther.

2014 Apr 2;16(2)

White D, Pahau H, Duggan E, Paul S, Thomas R. Trajectory of intensive

treat-to-target disease modifying drug regimen in an observational study of

an early rheumatoid arthritis cohort. BMJ Open. 2013 Jul 31;3(7).

Law SC, Street S, Yu CH, Capini C, Ramnoruth S, Nel HJ, van Gorp E,

Hyde C, Lau K, Pahau H, Purcell AW, Thomas R. T-cell autoreactivity to

citrullinated autoantigenic peptides in rheumatoid arthritis patients carrying

HLA-DRB1 shared epitope alleles. Arthritis Res Ther. 2012 May 17;14(3).

Conference abstracts

CIMT in individuals with Rheumatoid Arthritis compared to individuals

with Type 2 Diabetes Mellitus [abstract].

Helen Pahau1, Brian Haluska3, Leanne Short3, Vibeke Videm*2, and

Ranjeny Thomas*1 (Oral presentation at New Zealand Rheumatology

Conference, 2014)



Helen Pahau, Brian Haluska, Leanne Short, Vibeke Videm, and Ranjeny

Thomas

(Poster presentation at Princess Alexandra Hospital Symposium, 2014)



Helen Pahau, Brian Haluska, Leanne Short, Vibeke Videm, and Ranjeny

Thomas

(Poster presentation at Australia Rheumatology Conference, 2014)

Cardiovascular disease is increased prior to onset of rheumatoid arthritis but

not osteoarthritis: the population-based Nord-Trøndelag health study

(HUNT) [abstract]

v

Helen Pahau, Matthew A Brown, Sanjoy Paul, Ranjeny Thomas and

Vibeke Videm

(Poster presentation at the Australian Health & Medical Congress, Adelaide

2012)



(HUNT) [abstract].


Vibeke Videm

(Poster presentation at the American College of Rheumatology Conference,

2012)



(HUNT) [abstract].


Vibeke Videm (Poster presentation at the Vascular Conference, Phuket,

2012)



(HUNT) [abstract].


Vibeke Videm

(Oral presentation at the Princess Alexandra Hospital Symposium, Brisbane

2012)

Outcome of an intensive treat-to-target disease modifying drug regimen in

early rheumatoid arthritis within in the first month sets the response

trajectory for one year [abstract]. Douglas White, Helen Pahau, Emily

Duggan, Sanjoy Paul & Ranjeny Thomas. (Poster presentation at the

Australian Rheumatology Conference, 2011).

vi

Publications included in this thesis

Pahau H, Brown MA, Paul S, Thomas R, Videm V. Cardiovascular disease

is increase prior to onset of rheumatoid arthritis but not osteoarthritis: the

population-based Nord-Trondelag health study (HUNT). Arthritis Res Ther.

2014 Apr 2;16(2) – incorporated as Chapter 2 (complete article in

Appendices).

Contributor Statement of contribution

Author-Pahau H (Candidate) Wrote the paper (85%)

Statistical analysis of data (80%)

Author-Videm V Wrote and edited paper (10%)

Statistical analysis of data (15%)

Author-Paul S Statistical analysis of data (5%)

Author-Thomas R Wrote and edited paper (5%)

Contributions by others to the thesis

Chapter 3: Dr Sanjoy Paul Bio statistician Director School of Population

Health, The University of Queensland, assisted with statistical analysis.

Chapter 4: Dr Brian Haluska, Leanne Short, Dr Carly Jenkins, CIRCA, The

University of Queensland performed the ultrasound images

Chapter 5: HPLC on serum of the RA and healthy controls were

completed by DR Rowsan Ara.

Rowsan assisted with reviewing of patients medical records and writing the

methods.

Significant contributions to the thesis as a whole have been made by

Professor Vibeke Videm and Professor Ranjeny Thomas in the capacity of

vii

my PHD supervisor. This includes conceptual and statistical analysis,

writing and editing and advice on thesis content.

Statement of parts of the thesis submitted to qualify for the award of

another degree None

Acknowledgements

E rere ana ngā mihi huri haere i te tihi o maunga Hikuragi maringi ana

ki te awa ō Waiapu hei whārikihia ngā kōiwi ō āku mātua tīpuna o

Ngati Porou.

Ko tōku oranga he tāonga tuku iho mai rā nō. Ko tēnei te tino hā o tōku

tino rangatiratanga. Mai te timatanga o te pō, ki te ao mārama.

Kō au tēnei he kākano i ruia mai Rangiatea.

He tāonga tēnei hei koha ki ōku Tuakana ā Patricia TeoArani Wilson

rāua kō Ellamein Matariki Phillips me ōku tungane ā Moana Pahau

rāua kō Tapara Pahau.

My acknowledgements flow around the tip of my mountan Hikurangi, into the Waiapu river

which nuture the bones of my beloved parents and tipuna.

My being is a gift formed long ago. This is the breath of my existence. From the beginning

of time to today.

I am a seed sown from Rangiatea.

This treasure I gift to my sisters Trish and Ellamein and my brothers Moana and Tapara.

There is no doubt that this thesis would not be possible without the support,

encouragement and guidance of my supervisor Professor Ranjeny Thomas.

Thank you for believing in me.

My sincere gratitude to Professor Vibeke Videm who guided me through the

world of statistics and provided valuable advice and friendship. Dr Douglas

White who never faulted in believing I would be able to finish this thesis.

You are a great mentor and I will be forever grateful.

Helen B thank you for the last minute readings, the feedback, friendship and

unrelated PHD talks and great coffee. Soi I am so fortunate to have met you,

you provided me with laughter and joy and of course the final nudge at the

viii

end. To my colleagues of the Thomas Lab, thank you for allowing me to

share your journeys and for being a part of mine.

Thank you to Professor Don Mathieson, Georgina Paerata, Terry Ehau,

Professor Jim Mann, Dr Kirsten Coppell, Dr Rawiri Tipene-Leach, Dr Sally

Abel and all the dinosaurs (you know who you are), who were instrumental

in my career pathway.

I would like to acknowledge the financial support I received from the

University of Queensland – Research Scholarship and the Diamantina

Institute – Research Scholarship.

To my loving parents, who gave us unconditional love and support and

encouraged us to be whoever we wanted to be. I am forever grateful and

fortunate. You are the best parents in the world.

To Shane and Kewene, Aunty Ma, and our beautiful neighbours at 50

Somerset St, I am forever grateful for your support and help with

TeWhaikura.

Finally I acknowledge my sons Tapara and Hamuera, who have grown up to

be special young men, we are very proud of you both. My girl TeWhaikura,

who quietly reminded me that life was not only about my PHD, thank you

my girl for keeping me real.

To my husband Sam, always supportive, always encouraging, I love you.

ix

Keywords

rheumatoid arthritis, cardiovascular disease, carotid intima-media thickness, kynurenine

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 110322, Rheumatology and arthritis 50%

ANZSRC code: 110201, Cardiology (Cardiovascular Diseases), 50%

Fields of Research (FoR) Classification

FoR code: 1103, Medical and Health Science, 50%

FoR code: 1102, Cardiorespiratory medicine and haematology, 50%

Table of Contents

Chapter 1: Literature Review ........................................................................................................... 1

1.1 Introduction ........................................................................................................................... 1

1.2 Epidemiology ........................................................................................................................ 1

1.3 Aetiology ............................................................................................................................... 1

1.3.1 Genetic Factors ................................................................................................................... 2

1.3.2 Environmental factors ......................................................................................................... 2

1.4 Pathogenesis .......................................................................................................................... 3

1.5 Clinical Features .................................................................................................................... 4

1.5.1 Classification Criteria ......................................................................................................... 7

1.5.2 Diagnosis and disease activity .......................................................................................... 10

1.6 Rheumatoid Factor .............................................................................................................. 10

x

1.6.1 Anti- citrullinated protein antibodies ................................................................................ 10

1.6.2 CRP and ESR .................................................................................................................... 10

1.7 Disease Activity Scores ....................................................................................................... 11

1.8 Imaging ................................................................................................................................ 13

1.9 Treatment............................................................................................................................. 13

1.10 Early arthritis clinics and ‘treat to target’ regimens .............................................................. 15

1.10.1 Early arthritis clinics ....................................................................................................... 15

1.10.2 Treat to target regimens .................................................................................................. 15

1.11 Mortality and Morbidity in RA.............................................................................................. 16

1.12 Cardiovascular Disease .......................................................................................................... 17

1.12.1 Epidemiology .................................................................................................................. 17

1.13 Atherosclerosis ...................................................................................................................... 17

1.13.1 Risk Factors for atherosclerosis ...................................................................................... 19

1.13.2 CRP and atherosclerosis ................................................................................................. 20

1.14 Effects of Dyslipidaemia in CVD .......................................................................................... 20

1.14.1 Lipid modification and CVD risk ................................................................................... 22

1.15 Assessment of CVD in RA .................................................................................................... 23

1.16 Kynurenine pathway .............................................................................................................. 25

1.17 Kynurenine in RA .................................................................................................................. 26

1.18 Rationale for the study ........................................................................................................... 27

1.18.1 Aims ................................................................................................................................ 27

Chapter 2: Cardiovascular disease is increased prior to onset of rheumatoid arthritis but not

osteoarthritis: HUNT population-based Norwegian health surveys. ................................................. 29

2.1 Abstract ............................................................................................................................... 29

2.2 Introduction.............................................................................................................................. 30

2.3 Patients and methods ........................................................................................................... 31

2.4 Statistical methods ............................................................................................................... 33

2.5 Results ................................................................................................................................. 35

xi

2.6 Discussion ........................................................................................................................... 40

2.6.1 CV disease and incident RA ............................................................................................. 40

2.6.2 Risk factors for osteoarthritis............................................................................................ 42

2.6.3 Inflammation as a link between CV disease and incident RA.......................................... 43

2.7 Conclusions ......................................................................................................................... 44

Chapter 3: T2D patients have a higher risk of carotid-intima media thickness progression than RA

patients ............................................................................................................................................... 46

3.1 Background ......................................................................................................................... 46

3.2 Introduction: ........................................................................................................................ 47

3.3 Materials and Methods ........................................................................................................ 48

3.3.1 Participants ....................................................................................................................... 48

3.3.2 Cardiovascular risk factor measurement: ......................................................................... 49

3.3.3 Cardiovascular events ....................................................................................................... 49

3.4 Measurement of intima media thickness ............................................................................. 50

3.5 Statistical analysis ............................................................................................................... 51

3.6 Results ................................................................................................................................. 52

3.6.1 Sample characteristics ...................................................................................................... 52

3.7 Comparison of CIMT in RA and T2D patients ................................................................... 53

3.8 Influence of medication on CIMT in RA patients ............................................................... 54

3.9 CV events ............................................................................................................................ 55

3.10 Discussion .............................................................................................................................. 55

3.10.1 Differences in CV risk factors in RA and T2D .............................................................. 55

3.10.2 Comparison of CIMT in RA and T2D patients .............................................................. 56

3.10.3 Effect of anti-rheumatic drugs on atherosclerosis progression ....................................... 57

3.10.4 Advantages of CIMT as a measure of CVD ................................................................... 57

3.11 Limitations ............................................................................................................................. 58

3.12 Conclusion: ............................................................................................................................ 59

Chapter 4: Kynurenine and tryptophan concentrations and kynurenine/tryptophan ratio in rheumatoid

arthritis patients: A 10year follow-up study. ..................................................................................... 64

xii

4.1 Background ......................................................................................................................... 64

4.2 Introduction ......................................................................................................................... 66

4.3 Materials and methods: ....................................................................................................... 68

4.3.1 Study Population ............................................................................................................... 68

4.3.2 Baseline data: .................................................................................................................... 68

4.3.3 10-year follow-up data:..................................................................................................... 68

4.4 Measurement of Kynurenine ............................................................................................... 68

4.5 Statistical Analysis: ............................................................................................................. 69

4.7 KYN concentrations in RA patients with and without the CV and malignancy endpoints of the

study ............................................................................................................................................... 73

4.7.1 A) New CV event ............................................................................................................. 73

4.7.2 B) New Malignancy .......................................................................................................... 74

4.7.3 C) Death ............................................................................................................................ 74

4.8 Causes of death for the RA patients with KYN data:.......................................................... 75

4.8.1 Other variables (than KYN) related to death .................................................................... 75

4.9 Discussion ........................................................................................................................... 77

4.9.1 The relationship between KYN and CVD ........................................................................ 77

4.9.2 Malignancy and KYN ....................................................................................................... 78

4.9.3 Previous studies of predictors of mortality in RA. ........................................................... 78

4.10 KYN concentration level and mortality ................................................................................. 79

4.11 Limitations ............................................................................................................................. 79

4.12 Conclusion ............................................................................................................................. 80

Chapter 5: Final Discussion ........................................................................................................... 81

Future research ............................................................................................................................... 82

References .......................................................................................................................................... 84

Appendices......................................................................................................................................... 97

xiii

List of Figures

Figure 1-1: In the healthy joint (a) the thin synovial membrane lines the non-weight-bearing aspects

of the joint. In rheumatoid arthritis (b) the synovial membrane becomes hyperplastic and infiltrated

by chronic inflammatory cells. Ultimately it develops into 'pannus', which migrates onto and into

the articular cartilage and underlying bone. ........................... 3

Figure 1-2: Disease activity score 28 ................................................................................................

12

Figure13:Recommendations...............................................................................................................15

Figure 1-4: Ten recommendations on treating rheumatoid arthritis to target based on both evidence

andexpert opinion: ......................................................................................................... 16

Figure 1-5: Algorithm for treating RA to target. Indicated as separate threads are the main target

(remission and sustained remission) and the alternative target (low disease activity in

patients with long-term disease), but the approaches to attain the targets and sustain

them are essentially identical. Adaptation of therapy should usually be done by

performing control examinations with appropriate frequency and using composite

disease activity measures which comprise joint counts

....................................................... 17

Figure 1-6: Schematic diagram of the kynurenine pathway ................................................ 24

Figure 2-1: Proposed model for illustrating the relationship among RA, CV disease, inflammation,

atherosclerosis and predisposing factors

.............................................................................. 40

Figure 3-1: Relationship between mean CIMT at follow up and baseline in T2DM and RA ...........

56

Figure 3-2: Yearly rate of change in carotid intima-media thickness in diabetes and rheumatoid

arthritis patients .................................................................................................................... 57

Figure 3-3: Carotid intima-media thickness at follow-up depending on DMARD treatment. ...........

58

Figure 4-1: KYN concentrations in RA patients and healthy controls ...............................................

65

Figure 4-2: TRP concentrations in RA patients and healthy controls.................................................

66

Figure 4-3: KYN/TRP ratio of RA patients and controls ...................................................................

67 Figure 4-4: KYN concentrations in RA patients with and without a new CV event between 2003

and 2013. .............................................................................................................................. 68

xiv

Figure 4-5: KYN concentrations in RA patients with and without a new malignancy between 2003

and 2013. .............................................................................................................................. 69

Figure 4-6: KYN concentrations in RA patients dying or alive between 2003 and 2013. .................

70

xv

List of Tables

Table 1-1: Criteria for inclusion as extra-articular manifestations of RA

................................................ 4

Table 1-2: The 1987 revised criteria for the classification of RA by the ACR

...... .................................. 7

Table 1-3: The 2010 American College of Rheumatology / European

League Against Rheumatism Classification criteria for rheumatoid

arthritis........................................................................ 9

Table 1-4: DAS 28 .............................................................................. 13

Table 2-1: Characteristics of individuals without rheumatoid arthritis (RA)

or osteoarthritis (OA) at baseline ................................................ 33

Table 2-2: Effects of risk factors on incident RA and osteoarthritis, by

multivariate logistic regression .. 36

Table 3-1: Characteristics of individuals with Rheumatoid Arthritis (n=78)

and individuals with T2DM (n=212) 54

Table 3-2: Characteristics of RA patients (n=78) at baseline and follow up

and data on CV events in RA and T2D ................................ 55

Table 4-1: Characteristics of RA 64

Table 4-2: KYN concentrations at baseline by cause of death.................. 71

Table 4-3: Other variables (than KYN) related to death .......................... 72

Table 4-4: Model with death as an outcome ............................................ 73

xvi

Abbreviations

HCQ Hydrochloroquine

HDL High density lipoprotein

HLA Human leukocyte antigen

HUNT North Trondelag Health Study

ICA Internal carotid artery

IDO1 Indoleamine 2,3 dioxygenase 1

IDO2 Indoleamine 2,3 dioxygenase 2

IFN Interferon Gamma

IgG Immunoglobulin antibodies

IHD Ischemic heart disease

IMT Intima media thickness

KYN Kynurenine

LDL Low density lipoprotein

MCE Major coronary events

MCP Metacarpophalengeal

MHC Major Histocompatibility complex

MI Myocardial Infarction

MMPs Metalloproteinase’s

MRI Magnetic resonance imaging

MTP Metatarsalphalengeal

MTX Methotrexate

NSAIDs Non steroidal anti inflammatory

drugs

OA Osteoarthritis

PIP Proximal interphalangeal

RA Rheumatoid arthritis

SCORE Systemic coronary risk evaluation

SE Shared epitope

SLE Systemic lupus erythematosus

SMR Standard mortality ratio

xvii

SSZ Sulphasalazine

T2D Type 2 diabetes

TDO Tryptophan 2,3 dioxygenase

TNF Tumor necrosis factor

TRP Tryptophan

VLDL Very low density lipoprotein

1

Chapter 1: Literature Review

Rheumatoid Arthritis

1.1 Introduction

Rheumatoid Arthritis (RA) is a common systemic inflammatory disorder

primarily involving the joints. This arthritis is progressive and if

inadequately treated, significant functional disability can occur within 10

years [1]. The pattern of disease activity varies in patients, and often

includes periods of high disease activity interspersed with periods of low

disease activity. The precise aetiology of RA is unknown, however a

combination of environmental and genetic factors are implicated [2]. There

is a reduced median life expectancy of 7-10 years for people with RA and

premature cardiovascular disease (CVD) is primarily responsible [3].

Despite treatment advances, RA patients continue to have significantly

higher mortality and morbidity rates than the general population [4].

At present there is no cure for RA. Current goals of treatment include (1)

early and intensive management; (2) suppression of disease activity; (3)

limitation of disease progression and (4) adequate treatment of co-

morbidities. Continued research and subsequent advances in therapeutics

have recently enhanced effective treatment of this disease.

1.2 Epidemiology

The annual incidence of RA is reported to be approximately 0.03 percent

[5]. Although RA can affect any age group the peak age of onset is between

30 and 55 years [6]. Women are affected two to three times more often than

men [7]. There is currently an overall decline in the incidence of RA and

this is more evident in the female population [8]. The disease prevalence is

approximately 1 percent in Caucasians, but varies between ethnicities

including 0.1 percent (in rural Africans) and 56 percent (in Pima and

Chippewa Indians) [5].

1.3 Aetiology

The exact cause of RA is unknown however a number of factors are

implicated in the aetiology. Current evidence suggests that both genetic and

environmental factors are involved.

2

1.3.1 Genetic Factors

Increasing evidence suggests RA is highly heritable. There is greater

disease concordance in monozygotic twins (12-15%) than dizygotic twins

(3.5%) [9, 10]. Recent population health studies have also shown that

prevalence of RA is higher in first degree relatives of people with RA [11,

12]. These twin and first-degree relative studies suggest genetic factors have

a significant impact on susceptibility to RA. Estimated genetic contribution

to RA is thought to be as high as 50% to 60% [13, 14].

The best-established genetic association is with the human leukocyte

antigen (HLA) locus, in particular the HLA-DR B1 chains known as the

shared epitope (SE) [15]. The SE are known to be associated with RA

susceptibility [16], specifically susceptibility to Anti-citrullinated protein

antibody (ACPA) positive RA [17].

The presence of anti-citrullinated protein antibody (ACPA) measured by

anti-cyclic citrullinated antibodies (anti-CCP) is highly specific for RA [18,

19] and are reported to be a good predictor for the development of RA [20].

In addition to the major histocompatibility complex (MHC) alleles (HLA in

humans), there are non MHC alleles that are associated with RA risk and

also other autoimmune diseases [21]. For instance PAD 14 is associated

only with RA, whilst PTPN22 is linked to RA, systemic lupus

erythematosus, type 1 diabetes and other auto immune diseases [22-24].

1.3.2 Environmental factors

Smoking is the most established and studied environmental risk factor for

RA [25, 26]. There is a strong association with smoking and sera-positive

RA, [27, 28] with established interactions between the HLA-DR shared

epitope (SE) in ACPA positive RA patients [29]. This gene environment

interaction does appear to differ between ethnic groups. A cross-sectional

study involving African Americans with early RA, did not find an

association between smoking and ACPA-positive RA patients [30].

Furthermore RA cases in an Asian cohort demonstrated that the

combination of SE alleles and smoking was associated with risk of RA,

however ACPA status was not [31].

3

The association between HLA SE and smoking appears to hold irrespective

of the number of cigarettes smoked. A recent study showed that there is a

40% higher risk among ever smokers compared to never smokers [32],

indicating that light or heavy exposure increases risk of developing RA,

likely due to activation of the immune system against citrullinated protein

antigens [33].

1.4 Pathogenesis

Alteration in immune function leads to joint inflammation, degradation and

destruction in RA. CD4+ T cells are regarded as the protagonists in

initiating the inflammatory response in RA [34]. Numerous cytokines are

expressed in synovial tissue and an imbalance between pro and anti-

inflammatory cytokines induces chronic inflammation [35]. The inflamed

synovium contains various macrophage and fibroblast derived cytokines

that further support the activation of CD4+ Tcells. Tumor necrosis factor

(TNF-α) and interleukin-1 (IL-1β), are considered key pro-inflammatory

cytokines in the pathogenesis of RA [36, 37]. In vitro studies suggest that

IL-1β cause cartilage destruction by stimulating the release of matrix

metalloproteinases and other degradative products increases bone

reabsorption by stimulating osteoclast differentiation and activation [38].

Damage to articular cartilage in RA begins at the cartilage pannus interface,

with progressive erosions occurring into subchondral bone [38].

4

Strand et al. Nature Reviews Drug Discovery 6, 75–92 (January 2007) | doi:10.1038/nrd2196

Figure 1-1: In the healthy joint (a) the thin synovial membrane lines the non-weight-

bearing aspects of the joint. In rheumatoid arthritis (b) the synovial membrane

becomes hyperplastic and infiltrated by chronic inflammatory cells. Ultimately it

develops into 'pannus', which migrates onto and into the articular cartilage and

underlying bone.

1.5 Clinical Features

The symptoms of RA predominantly consist of pain, swelling and stiffness

of the joints. RA typically begins insidiously, with the signs and symptoms

developing slowly over weeks to months. However in a third of the patients,

onset is rapid and can occur over days or weeks [39].

The metacarpophalangeal (MCP), proximal interphalangeal (PIP),

metatarsalphalangeal joints (MTP) and the wrist are the most common joints

affected. In some patients the shoulders, elbows, knees and ankles may also

be affected. If this disease is left untreated, deformity of the joint and

disability can occur within 10 years of onset of disease [40]. However

achieving remission is within reach, and the earlier that RA treatment is

started the better the odds are of reaching remission.

5

RA patients with extra articular manifestations have increased morbidity

and mortality [41] and extra articular manifestations appear to be most

common in patients with severe disease activity [42]. Severe weight loss,

osteoporosis and CVD are examples of extra articular manifestations [43].

Table 1-1: Criteria for inclusion as extra-articular manifestations of RA

Criteria for inclusion as extra-articular manifestations of RA

1. Pericarditis (A) Clinical judgment and exudation verified by echocardiography

If ultrasound not available: criteria according to Hara et al4

(B) Clinical criteria: (1 required)

Typical pericardial pain, peripheral oedema, dyspnoea/orthopnoea, ascites, dysrythmia (heart rate >140/min, atrial flutter/fibrillation, 2–3 degree atrioventricular block, ventricular tachycardia)

Objective criteria compatible with pericarditis: (1 required)

Physical examination

Cardiac catheterisation findings

Histological examination

Other causes improbable, such as tuberculosis or other infection, metastases, primary tumour, postoperative status or other trauma

2. Pleuritis Clinical judgment and exudation verified by x ray examination

Other causes improbable, such as tuberculosis or other infection, metastases, primary tumour, postoperative status or other trauma

3. Felty’s syndrome Splenomegaly (clinically evident or measured by ultrasound) and neutropenia (<1.8×109/l) on two occasions

Other causes improbable, such as drug side effect or infection

4. Major cutaneous vasculitis Diagnostic biopsy or clinical judgment by dermatologist

5. Neuropathy

Clinical judgment by doctor and signs of polyneuropathy/mononeuropathy at electromyography/electroneurography

Other causes, such as diabetes or alcohol abuse, improbable

6. Scleritis, episcleritis or retinal vasculitis Clinical judgment by ophthalmologist

7. Glomerulonephritis Clinical judgment by nephrologist and positive biopsy

8. Vasculitis affecting other organs

Clinical judgment by organ specialist and biopsy compatible with vasculitis

9. Amyloidosis Clinical judgment and positive biopsy from affected organ

10. Keratoconjunctivitis sicca Clinical judgment

Positive rose-bengal staining or result of Schirmer’s test <5 mm/min

11. Xerostomia Clinical judgment

Abnormal sialometry, sialography, salivary scintigraphy or salivary gland biopsy with lymphocytic infiltrate

12. Secondary Sjögren’s syndrome Two of three criteria:

Keratoconjunctivitis sicca (see above)

Xerostomia (see above)

6

Serological evidence: rheumatoid factor, ANA, anti-Ro (SSA), anti-La (SSB) positive, or hypergammaglobulinaemia

13. Pulmonary fibrosis Clinical judgment and

Decreased vital capacity or carbon dioxide transfer factor by 15% from normal

14. Bronchiolitis obliterans organising pneumonia Clinical judgment by pulmonologist

15. Cervical myelopathy Clinical judgment

Increased atlantoaxial movement—verified by x ray examination

16. Subcutaneous rheumatoid nodules Clinical judgment

17. Rheumatoid nodules in other locations Positive biopsy

Downloaded from http://ard.bmj.com/ on December 17, 2014 - Published by

group.bmj.com

7

1.5.1 Classification Criteria

The 1987 American College of Rheumatology (ACR) criteria were used to

classify patients with RA until 2010 (Table 2). For the diagnosis of RA,

patients must fulfil 4 out of 7 criteria including morning stiffness 1 hour,

arthritis of 3 joint areas, arthritis of hand/wrist joints, symmetrical arthritis,

rheumatic nodules, serum rheumatoid factor (RF) and radiographic changes

[44]. The criteria distinguished patients with established RA from other

arthritides and identified patients with moderate to severe disease.

Table 1-2: The 1987 revised criteria for the classification of RA by the ACR

Criterion Description

1. Morning Stiffness Morning stiffness in

and around the joints,

lasting at least one hour before

maximal improvement

2. Arthritis of 3 or more joint

areas At least three joint areas (out

of 14 possible; right or left

PIP, MCP, wrist, elbow,

ankle, MTP joints)

simultaneously have had soft

tissue swelling or fluid (not

bony overgrowth) as observed

by a physician

3. Arthritis of hand

joints At least one area swollen (as

defined in criterion two) in a

wrist, MCP, or PIP joint

4. Symmetric arthritis Simultaneous involvement (as

in criterion two) of the same

joint areas on both sides of the

body (bilateral involvement of

PIP, MCP, or MTP joints

without absolute symmetry is

acceptable)

8

5. Rheumatoid nodules Subcutaneous nodules over

bony prominences or extensor

surfaces, or in juxta-articular

regions as observed by a

physician

6. Serum rheumatoid

factor Demonstration of abnormal

amounts of serum rheumatoid

factor by any method for

which the result has been

positive in <5% of normal

control subjects

7. Radiographic changes Radiographic changes typical

of rheumatoid arthritis on

posteroanterior hand and wrist

radiographs, which must

include erosions or

unequivocal body

decalcification localized in, or

most marked adjacent to, the

involved joints (osteoarthritis

changes alone do not qualify)

In 2010 the American College of Rheumatology (ACR) and the European

League Against Rheumatism (EULAR) revised the classification criteria to

allow capture of RA patients with early and milder disease, while still

allowing application to patients with established and more severe disease

(Table 3) [45].

9

Table 1-3: The 2010 American College of Rheumatology / European League Against Rheumatism classification criteria for

rheumatoid arthritis

Score

Target population (Who should be tested?): Patients who

1) have at least 1 joint with definite clinical synovitis (swelling)

2) with the synovitis not better explained by another disease

Classification criteria for RA (score-based algorithm: add score of categories

A–D;

a score of ≥6/10 is needed for classification of a patient as having definite

RA

A. Joint involvement

1 large joint

2-10 large joints

1-3 small joints (with or without involvement of large joints)

4-10 small joints (with or without involvement of large joints)

>10 joints (at least 1 small joint)

B. Serology (at least 1 test result is needed for classification)

Negative RF and negative ACPA

Low-positive RF or low-positive ACPA

High-positive RF or high-positive ACPA

C. Acute-phase reactants (at least 1 test result is needed for classification)

Normal CRP and normal ESR

Abnormal CRP or normal ESR

D. Duration of symptoms <6 weeks

≥6 weeks

0

1

2

3

5

0

2

3

0

1

0

1

10

1.5.2 Diagnosis and disease activity

The diagnosis of RA is made via a thorough assessment of patient

symptoms, physical examination and additional investigations in accordance

with the classification criteria.

1.6 Rheumatoid Factor

Rheumatoid factor (RF) is a complex of antibodies directed against the

conserved end of immunoglobulin (IgG) antibodies. It occurs in 60% of

patients at diagnosis and rises to approximately 80% with time. It is

associated with more severe disease and a greater risk of atherosclerosis.

The specificity for RF is however relatively low (60-70%) and is also

detected in patients with other diseases and also in the healthy population.

1.6.1 Anti- citrullinated protein antibodies

ACPA are auto-antibodies directed against peptides and proteins that are

citrullinated. ACPA are used as a clinical biomarker for the diagnosis of

RA, and have a 55-80% sensitivity and specificity of 90-98% for RA [46,

47]. The presence of ACPA is also a prognostic marker and is associated

with more aggressive disease [48], preceding the clinical manifestations of

RA by up to 10 years along with RF [19, 49]. Furthermore, there is evidence

of significant interactions between the HLADRB1 shared epitope and

cigarette smoking in the development of ACPA-positive RA [50, 51]. Sero-

positive RA is more common (70%) than sero-negative disease (30%) and is

associated with more aggressive disease and extra- articular manifestations

[52].

1.6.2 CRP and ESR

C - reactive protein (CRP) is a member of the pentraxin family, a pattern-

recognition protein and a host-defence-related component of the innate

immune system [53-55]. CRP was discovered in 1929 and is the most

characteristic acute phase protein in the human body [56]. CRP is a sensitive

marker of systemic inflammation, and the plasma concentration levels are

elevated in RA patients and in other conditions with both acute and chronic

inflammation [56-58]. Although CRP is a non-specific test, doctors may

11

utilise the test to assist the effectiveness of a specific arthritis treatment and

to monitor periods of disease flare-up.

Erythrocyte sedimentation rate (ESR) measures how fast red blood cells fall

through a column of blood. It is an indirect index of acute-phase protein

concentrations (particularly fibrinogen) and is a sensitive but non-specific

test used to assist in the detection of inflammation associated with auto

immune diseases, infections and cancers [59]. Similar to CRP, doctors may

utilise ESR to test the effectiveness of specific arthritis treatment and to

follow inflammation levels.

1.7 Disease Activity Scores

In order to evaluate the need for, and the effect of treatment in patients with

RA, it is valuable to measure the disease activity at diagnosis and follow-up

in a standardized manner. For this purpose several scores have been

developed. The Disease Activity Score (DAS) was developed in Europe as a

continuous activity score [60]. This score takes into account the patient’s

general health, the number of swollen and tender joints as well as the ESR: a

non-specific measure of inflammation. The DAS28 takes into account 28

different joints. A value >5.1 is regarded as high disease activity, <3.2 is

regarded as low activity and <2.6 is regarded as remission. Figure 2 by

comparing a patient’s DAS28 score over multiple time points, one can

substantiate his/her improvement or response. The EULAR response criteria

are shown in Figure 3.

12

Figure 1-2: Disease activity score 28

From: DAS-Score.nl. Available at http://www.das-score.nl/www.das-

score.nl/index.html

Table 1-4: DAS 28

PRESENT DAS28 DAS28 IMPROVEMENT OVER TIME POINTS

>1.2 0.6-1.2 <0.6

<3.2 Good response Moderate response No response

3.2-5.1 Moderate response Moderate response No response

>5.1 Moderate response No response No response

DAS Score

DAS28 >5.1

= high disease activity

DAS28 <3.2

DAS28 <2.6 = remission

= low disease activity

Treatment response over time is also a crucial component of patient

assessment, although largely a tool for clinical trial ACR score measures %

treatment response, i.e., ACR20 requires at least a 20 percent improvement

in both the number of swollen and tender joints, as well as a 20 percent

13

improvement in three of the following five parameters: ESR, health

assessment questionnaire (HAQ), physician’s global assessment of disease

activity, patient’s global assessment and patient’s pain [61].

Correspondingly, the ACR50 requires a 50% improvement and the ACR70

a 70% improvement. Furthermore, the ACR has defined remission criteria

based on the same variables [62].

1.8 Imaging

Radiography remains the mainstay of imaging in the diagnosis and follow-

up of RA. RA patients should have a hand and wrist x-ray, and baseline x-

rays of affected joints to document any future erosive changes. Erosions

often develop in the first year but may occur at any time. A previous study

showed radiographically demonstrable erosions were present in 30% of their

patients at diagnosis, and in 70% 3 years later [63]. The use of newer

imaging modalities such as ultrasonography, magnetic resonance imaging

(MRI) and computerized tomography (CT) are increasingly being used to

assist in the diagnosis and continuing management of RA [64]. Many of the

imaging methods offer greater sensitivity in detecting synovitis, effusions,

joint subluxations, joint space narrowing and bone edema then standard x-

ray [65]. Furthermore, when compared with conventional methods, the

increasing use of ultrasound has allowed for a greater success rate when

aspirating joints, and greater accuracy when injecting steroids into affected

joints [65].

1.9 Treatment

RA rarely spontaneously remits. As the exact pathophysiology is unknown

and there is currently no cure, treatment aims to suppress and modulate the

inflammatory disease. Supportive ancillary treatments such as

physiotherapy, occupational therapy, podiatry and surgery are used as

adjuncts. Medications used in RA aim to suppress the ‘overactive’ immune

system. Different drugs with various mechanisms of action are utilised and

often a combination of therapies is needed.

Non-steroidal anti-inflammatory agents (NSAIDs) have analgesic and anti-

inflammatory effects. They are used effectively to assist with RA

14

symptoms. They also have significant gastric and cardiac toxicity [66]. The

gastrointestinal toxicity appears to be related to cyclooxygenase 1 (COX)

inhibition. However COX-2 inhibitors with an improved gastrotoxicity

profile have been found to increase the risk of atherothrombosis even with

short-term use.

Current medications used to treat RA span the history of medicine, from

intramuscular gold to targeted cytokine blocking agents. The exact

mechanism of action of the many of the disease modifying agents

(DMARDs) is incompletely understood. Disease modifying (or slow

acting) antirheumatic drugs include Methotrexate (MTX), Sulphasalazine

(SSZ), Leflunomide, Hydroxchloroquine (HCQ), Gold, Cyclosporine,

Azathioprine and Corticosteroids. Biologic (B)DMARDs are newer

therapeutics and they interfere with the immune system in a more selective

way, targeting cytokines or signalling pathways. However, both DMARDs

and B-DMARDs suppress immune and inflammatory pathways and hence

RA activity.

Methotrexate is the most frequently used DMARD in RA and the

foundation therapeutic for DMARD and biologic combinations [67]. Results

of multiple double-blind, randomised studies have indicated that MTX

improves the health of patients (e.g. improving health assessment

questionnaire scores and those of other quality of life measures), and

decreases systemic inflammation (e.g. erythrocyte sedimentation rate,

concentrations of CRP and signs of clinical inflammation [68, 69].

Patients react differently to anti-rheumatic medications and often require

varying doses or combinations to manage their disease. Adverse effects are

not uncommon. The medications used to treat RA mostly take several

months or longer to achieve remission. However over the last decade, the

development of new and effective treatments and treatment regimens has

substantially improved the prognosis for patients with RA. Many patients

now experience lower disease activity and remission than in the 1970s and

1980s [70].

15

1.10 Early arthritis clinics and ‘treat to target’ regimens

1.10.1 Early arthritis clinics

There is substantial evidence that early treatment with DMARDs can lead to

better outcomes, including the suppression of synovitis, prevention of bone

destruction, and the reduction of functional disability and mortality rates.

The late 80s saw the introduction of the concept of ‘early intervention’ [71].

Soon after, specialised early arthritis clinics (EAC) [72] were created in

multiple centres around the world, ensuring early diagnosis and treatment.

The EAC most often represents a structured environment where patients can

be assessed, reviewed and treated in a protocol-driven manner, utilising

best-practice therapeutics and technology.

1.10.2 Treat to target regimens

In 2008, an international expert panel primarily of Rheumatologists

reviewed available goal directed therapy in RA, resulting in the

International Treat to Target initiative [73]. The panel introduced

overarching principles (Figure 3) and 10 recommendations for “Treat

Arthritis to Target” (Figure 4). There are extended guidelines from both

European and American Rheumatology expert groups which, although

generally similar, vary in the type of medications used. The aim of the treat-

to-target therapeutic approach is to bring RA under control quickly and to

prevent joint damage and disease progression.

Overarching principles:

a. The treatment of rheumatoid arthritis must be based on a shared decision between patient and rheumatologist.

b. The primary goal of treating the patient with rheumatoid arthritis is to maximize long-term, health-related

quality of life through control of symptoms, prevention of structural damage, normalization of function and social

participation.

c. Abrogation of inflammation is the most important way to achieve these goals.

d. Treatment to target by measuring disease activity and adjusting therapy accordingly optimizes outcomes in

rheumatoid arthritis.

Figure 1-3: Recommendations

16

1. The primary target for treatment of rheumatoid arthritis should be a state of clinical remission. 2. Clinical remission is defined as the absence of signs and symptoms of significant inflammatory disease

activity.

3. While remission should be a clear target, based on available evidence low disease activity may be an

acceptable alternative therapeutic goal, particularly in established long-standing disease.

4. Until the desired treatment target is reached, drug therapy should be adjusted at least every 3 months.

5. Measures of disease activity must be obtained and documented regularly, as frequently as monthly for patients

with high/moderate disease activity or less frequently (such as every 3–6 months) for patients in sustained low

disease activity or remission.

6. The use of validated composite measures of disease activity, which include joint assessments, is needed in

routine clinical practice to guide treatment decisions.

7. Structural changes and functional impairment should be considered when making clinical decisions, in

addition to assessing composite measures of disease activity.

8. The desired treatment target should be maintained throughout the remaining course of the disease.

9. The choice of the (composite) measure of disease activity and the level of the target value may be influenced

by consideration of co-morbidities, patient factors, and drug-related risks.

10. The patient has to be appropriately informed about the treatment target and the strategy planned to reach this

target under the supervision of the rheumatologist

Figure 1-4: Ten recommendations on treating rheumatoid arthritis to target based on both evidence and

expert opinion: From: Ann Dis Rheum. 2010, 69:631-637 [73]

An algorithm was proposed for treating RA to target (Fig 4) and is based on the recommendation in

Fig 3.

Figure 1-5: Algorithm for treating RA to target. Indicated as separate threads are the main

target (remission and sustained remission) and the alternative target (low disease activity in

patients with long-term disease), but the approaches to attain the targets and sustain them are

essentially identical. Adaptation of therapy should usually be done by performing control

examinations with appropriate frequency and using composite disease activity measures which

comprise joint counts

From: Ann Rheum Dis. 2010;69:631-637 [73]

1.11 Mortality and Morbidity in RA

Patients with RA have increased mortality and morbidity when compared to

healthy controls [74, 75]. The increased mortality predominantly relates to

increased cardiovascular disease [76-79]. It was found in a retrospective

study of an inception cohort of RA patients that they were at increased risk

of cardiovascular death, ischemic heart disease and heart failure compared

17

to age and sex matched community controls. The incidence of

cardiovascular events was significantly higher among those with RA

compared with controls (3.43 versus 0.59 per 100 patient-years) [80].

Another study of 3501 patients with RA with follow-up for 35 years, found

that mortality was increased twofold, resulting in a decreased lifespan of 7

to 10 years [81]. Mortality directly due to RA itself was low (9.8 percent of

the deaths). Other studies suggest that increased inflammation associated

with RA appears to contribute substantially to increased cardiovascular

mortality [82].

1.12 Cardiovascular Disease

CVD is the result of damage and dysfunction within the vascular tree. This

damage occurs as a result of loss of vessel compliance, damage to

components of vessel walls leading to plaques, evolution of aneurysms and

exposure of thrombotic wall components with resultant clot formation.

This damage causes the vascular tree to fail in different ways, leading to

particular clinical states.

1.12.1 Epidemiology

CVD is common in the general population, affecting the majority of adults

over the age of 60 years. The prevalence of coronary heart disease (CHD) is

approximately 1/3 to 1/2 that of total CVD. The lifetime risk of CHD was

illustrated in a study of 7733 participants, age 40 to 94, in the Framingham

Heart Study who were initially free of CHD [83]. The lifetime risk for

individuals at age 40 was 49% in men and 32% in women. Even those who

were free from disease at age 70 had a lifetime risk of 35% in men and 24%

in women [83]. While young patients in a large autopsy study had fatty

streaks [84], not all adults develop clinically apparent CVD as fatty streaks

progress at different rates in different individuals [85].

1.13 Atherosclerosis

Generalised thickening of the arteries with subsequent loss of wall

compliance is known as arteriosclerosis. Atherosclerotic heart disease,

otherwise known as coronary heart disease or coronary artery disease, is

most often caused by atherosclerosis that occurs due to the deposition of

18

cholesterol in the arterial wall causing plaque accumulation. Resulting

chronic inflammation in the plaques can lead to the thinning and rupture of

the endothelial cap and exposure of thrombogenic core. This then leads to

the formation of a platelet thrombosis often leading to vessel occlusion and

tissue infarction [86]. The presence of atherosclerosis is associated with

arteriosclerosis.

The vascular endothelium was previously considered to be an inert barrier

between the blood and tissues, however increasing evidence suggests that

the endothelium is an active biological interface. The single layer of

continuous endothelium lining arteries and veins forms a unique

thromboresistant layer between blood and potentially thrombogenic

subendothelial tissues [87, 88].

The exact pathogenesis of endothelial damage is not known, however a

range of risk factors has been shown to be associated with accelerated

vascular damage. These include environmental and genetic factors such as

smoking, hypertension, obesity, ageing, lack of exercise, inflammation,

diabetes and family history. Intima media thickness (IMT) increase begins

with endothelial dysfunction and is generally considered the earliest sign of

atherosclerosis.

The activity of endothelial cells includes regulation of vascular tone and

structure. This is accomplished by releasing molecules such as nitric oxide

and prostacyclin, which control vasodilatation and vasoconstriction [89].

Endothelial cells also provide both a non-thrombotic and non-adherent

arterial surface for leukocytes [88]. Endothelial dysfunction therefore is

defined as impaired vasodilatation or vasoconstriction to stimuli, pro-

inflammatory surface state, and prothrombotic surface state [90].

The changes usually occur slowly over time and vary greatly from

individual to individual. The first observable change of atherosclerosis is

the formation of a fatty streak. In one autopsy study in the US, for example,

all 2876 men and women, aged 15 to 34 years, had aortic fatty streaks [91].

The fatty streak is caused by an excess deposition of lipid in macrophages in

the vessel wall. With time these cells swell, there is deposition of

19

extracellular lipid and the plaque grows. Fibrous tissue is deposited, areas

calcify and the plaque increases in size. The increasing plaque size can

encroach on the vessel lumen, reducing blood flow and causing ischaemia

of downstream tissue, and causing symptoms such as angina. With time, the

fibrous cap thins, and with exposure to the shear stress of the turbulent

blood flow, is at risk of rupture. Upon rupture, the thrombogenic contents

rapidly cause clotting leading to vessel occlusion. This occlusion can lead

to tissue death, such as a myocardial infarction (MI) [92].

Concurrent with the plaque formation there is generalised vascular damage

where the lining of the vessel thickens. There is good correlation between

generalised arteriosclerosis, focal atherosclerosis and risk of CVD [93].

This arteriosclerosis causes a loss of vessel wall compliance, which

increases the load on the vessel and the heart. In healthy vascular trees this

is a reflection of the normal systolic outflow. The wave of ejected blood

from the heart is partially reflected back from the peripheral arteries.

Normally this is reflected back in diastole and does not load the heart.

However as the vessels lose compliance and stiffen this wave is reflected

back earlier and eventually arrives in systole. This reflected wave adds

additional load to the heart and the extra workload accelerates the

progression of heart disease [94]. Decreased aortic compliance has been

shown to be predictive of primary coronary events in hypertensive patients

with normal renal function [95].

1.13.1 Risk Factors for atherosclerosis

There are many risk factors for atherosclerosis. These factors show

individual variation and no single factor acts in isolation. The worldwide

INTERHEART study of patients from 52 countries, identified nine

potentially modifiable factors accounting for over 90 percent of the

population with an attributable risk of a first MI [96]. These included

smoking, dyslipidaemia, hypertension, diabetes, abdominal obesity,

psychosocial factors, regular alcohol consumption, not consuming a daily

intake of fruits and vegetables, and not regularly participating in physical

activity. Non modifiable risk factors include male sex, age, and a family

history of CVD.

20

1.13.2 CRP and atherosclerosis

The association of elevated CRP concentrations and increased risk of CVD

is well documented [97]. CRP levels are elevated in response to

inflammatory stimuli; therefore it is thought that CRP may play a direct role

in the development of atherosclerosis in inflammatory disease such as RA

[98, 99].

1.14 Effects of Dyslipidaemia in CVD

Lipid abnormalities play a critical role in the development of

atherosclerosis. Lipids such as cholesterol and triglycerides are insoluble in

plasma. Circulating lipid is carried in lipoproteins that transport the lipid to

various tissues for energy utilisation, lipid deposition, steroid hormone

production, and bile acid formation [100]. Lipoprotein consists of esterified

and unesterified cholesterol, triglycerides, and phospholipids and protein

[100]. The protein components of the lipoprotein are known as

apolipoproteins. The different apolipoproteins serve as cofactors for

enzymes and ligands for receptors. There are five major lipoproteins, each

of which has a different function. In clinical practice measurement of the

low density lipoprotein (LDL), high density lipoprotein (HDL) and total

cholesterol level are the most relevant [101].

Animal studies have demonstrated accelerated atherosclerosis with a high

cholesterol diet [102]. In humans, worldwide epidemiologic studies show an

increasing incidence of atherosclerosis when serum cholesterol

concentrations are > 3.9 mmol/L. The different lipoproteins have particular

effects on vessel health. In the clinical context high levels of LDL and low

levels of HDL are particularly important risk factors for CVD [103].

LDL cholesterol can promote inflammatory and immune changes via

cytokine release from macrophages and antibody production. Circulating

LDL rapidly accumulates in the cholesterol enriched macrophages (foam

cells), but not directly in the lipid core of atherosclerotic plaque [104].

Oxidative modification of LDL is necessary for macrophage uptake via

unregulated scavenger receptors and accelerated accumulation of cholesterol

[105]. Macrophage uptake of LDL cholesterol may initially be an adaptive

21

response, which prevents LDL-induced endothelial injury [106].

Cholesterol accumulation in foam cells leads to mitochondrial dysfunction,

apoptosis, and necrosis with resultant release of cellular proteases,

inflammatory cytokines, and prothrombotic molecules [106].

Oxidized LDL can cause disruption of the endothelial cell surface [107],

promote inflammatory and immune changes via cytokine release from

macrophages and antibody production, and increase platelet aggregation. It

may also play a role in plaque instability. Levels of oxidized LDL are

increased in patients with an acute coronary syndrome and are positively

correlated with the severity of the syndrome [108]. Low HDL levels are

more common in patients with a first myocardial infarction than in age-

matched controls without CHD (19 versus 4 percent) [109].

HDL, in contrast to LDL, has antiatherogenic properties that include reverse

cholesterol transport, maintenance of endothelial function, protection

against thrombosis, and maintenance of low blood viscosity through a

permissive action on red cell deformability [110]. The net effect is that

there is an inverse relationship between plasma HDL-cholesterol levels and

cardiovascular risk. Values above 1.9 mmol/L are associated with a

longevity syndrome [111].

Much of the anti-atherogenic effect of HDL is thought to be mediated by

reverse cholesterol transport, a process whereby excess cholesterol in cells

and in atherosclerotic plaques is removed [112]. Initially hepatic and

intestinal cells synthesise small nascent HDL particles composed of

phospholipids and apolipoproteins [113]. These HDL particles gather

surface components (phospholipids, cholesterol, and apolipoproteins) from

triglyceride-depleted chylomicron and very low density lipoprotein (VLDL)

remnants [113]. They also acquire free (unesterified) cholesterol from tissue

sites and other lipoproteins, as the initial HDL particles contain relatively

little cholesterol. Apo A-I on the surface of HDL plays a central role in this

process. It serves as a signal transduction protein to mobilize cholesterol

esters from intracellular pools [113].

22

1.14.1 Lipid modification and CVD risk

There is extensive evidence that modification of dyslipidaemia can lower

the risk of CVD. This evidence is both for primary (patients without

evidence of CVD) and secondary (patients with CVD) prevention. There

are multiple studies which assess the relationship between dyslipidaemia

and CVD. A meta-analysis of 38 primary and secondary prevention trials

found that for every 10 percent reduction in serum cholesterol, coronary

heart disease mortality was reduced by 15 percent and total mortality risk by

11 percent. No increase in non-coronary heart disease mortality was seen

[114, 115].

Lipid modification is achieved by medications, predominantly by HMG-

CoA reductase inhibitors also known as statins. They lower cholesterol by

inhibiting the enzyme HMG-CoA reductase, which is the rate-limiting

enzyme of the mevalonate pathway of cholesterol synthesis [116]. Inhibition

of this enzyme in the liver stimulates LDL receptors, resulting in an

increased clearance of LDL from the bloodstream and a decrease in blood

cholesterol levels. The first results can be seen after one week of use and the

effect is maximal after four to six weeks [116]. However the mechanisms by

which lipid-lowering therapy (particularly with statins) is beneficial, are

incompletely explained by the serum LDL concentration at baseline or after

treatment [117]. Although statins probably cause regression of

atherosclerosis, an improvement in outcome can be demonstrated as early as

six months: a time considered too early for significant plaque regression

[118]. Among the non-lipid mechanisms that may be involved in CVD

reduction are plaque stabilization, reduced inflammation, reversal of

endothelial dysfunction, and decreased thrombogenicity [119].

One important effect of statin therapy may be maintenance of integrity of

the fibrous cap of the plaque, thereby protecting against plaque rupture. This

effect appears to be mediated by inhibition of macrophage proliferation,

reduced expression of matrix metalloproteinases (MMPs) and tissue factor

(which promotes thrombus formation) by macrophages, and an increase in

tissue inhibitor of metalloproteinase-1 [120]. Statin therapy, given as

primary or secondary prevention, reduces the serum CRP concentration, an

23

effect that is mostly unrelated to lipid levels at baseline or during therapy

[121]. The fall in serum CRP begins within 14 days of treatment [122].

Additional evidence that statins may have an anti-inflammatory effect is

provided by a randomized trial that found that patients with rheumatoid

arthritis experienced modest clinical improvement in their RA with

atorvastatin. Atorvastatin also reduced CRP levels and the ESR compared

with placebo [123]. In clinical trials, statins appear to have greater effects in

patients with evidence of inflammation at baseline. The potential

importance of statin-induced reduction in serum markers of inflammation

was illustrated by an analysis from the secondary prevention cholesterol and

recurrent event (CARE) trial [124]. Patients with baseline serum

concentrations of CRP and serum amyloid A in the highest quintile had a

relative risk for a recurrent event 75 percent higher than those with levels in

the lowest quintile.

How statins might interfere with the inflammatory response is not well

understood. One possible mechanism is impairment of inflammatory cell

adhesion by inhibition of the main beta-2 integrin, LFA-1[125, 126]. Other

contributing factors may include reduced lipidation of intracellular proteins

and reduced expression of major histocompatibility complex class II

molecules on antigen presenting cells in response to interferon, decreasing

subsequent T-lymphocyte activation [127].

1.15 Assessment of CVD in RA

It is hard to power interventional studies in RA to look at definitive CVD

endpoints. This is because it is difficult to obtain a large enough cohort of

patients to show a treatment difference in mortality or cardiovascular events.

Therefore a surrogate measure of a person’s risk for future CVD is ideal. In

recent years, the evolution of non-invasive ultrasound techniques has

allowed assessment of cardiovascular health. These techniques can detect

vascular dysfunction, and appear to correlate well with future risk of CVD.

An ideal situation to assess a person’s risk for CVD would be to know the

exact state of their vascular tree. Currently there is no such technique. All

current tests are necessarily approximate measures of vascular risk.

Individual tests assess various aspects of vascular structure and function to

24

add weight to the knowledge of a persons’ CVD risk. No single test gives a

complete assessment of CVD.

Cardiac catheter studies can demonstrate the vessel lumen, but these tests

are expensive, invasive, have inherent risks and thus make them less useful

for large scale progress studies. Other tests, such as CT assessment of

coronary calcium scores, have good correlation with CVD but are also

expensive and expose the patient to significant levels of radiation.

Functional tests identify disease in the setting of significant coronary

stenosis, which implies the presence of advanced disease.

Ultrasound provides a range of non-invasive techniques that assess

peripheral vascular structure and function. It is commonly used to assess

arterial structure and function by measuring arterial wall thickness, arterial

wall reactivity and arterial stiffness. Observational studies have shown that

these tests correlate well with future risk of CVD and current CVD burden.

Atherosclerosis is a patchy process and it is difficult to gain an overall

impression of the burden of disease. It has been recognised that there is

good correlation between the risk of CVD and the thickness of the arterial

wall. This thickness can be measured using ultrasound.

Arteries are divided into three layers: the intima, media and adventitia.

Ultrasound detects changes in tissue boundaries and as such can clearly

delineate the discrete areas of the arterial wall. This allows the thickness of

the intima media to be measured. Measurement of the carotid artery is

preferred because it allows clear visualisation of a large calibre artery that is

relatively superficial, parallel and has a distinct anatomical area, the carotid

bulb, to provide a reference point, thus giving the most reliable data [128].

Carotid ultrasonography has been used to obtain measurements of the

thickness of the intima and media of the carotid arteries. Previous studies

have shown cross-sectional associations between common-carotid-artery

intima-media thickness and cardiovascular risk factors [129-131], the

prevalence of cardiovascular disease, [130-132] and the involvement of

other arterial beds with atherosclerosis [133] Changes in common-carotid-

25

artery intima-media thickness (CIMT) have also been adopted as a surrogate

end point to determine the success of interventions that lower the levels of

low-density lipoprotein cholesterol [134, 135]. CIMT has been shown to be

associated with diabetes, fibrinogen level, body mass index, and clinically

overt atherosclerosis [136]. There is evidence that CIMT can predict an

increased risk of future myocardial infarction and stroke.

Increasing levels of CIMT are associated with an incremental risk of

myocardial infarction and stoke. In one study, 5858 patients >65 years of

age with no pre-existing cardiovascular disease were separated into quintiles

on the basis of maximal CIMT, and followed for a mean period of 6.2 years.

Over this period, 47 of 897 patients in the lowest quintile of IMT (<0.87

mm) had a myocardial infarction or stroke. In comparison, those in the

highest quintile of IMT ( 1.18 mm) had 167 events. This equated to an

age- and sex-adjusted relative risk of 2.85 [93].

In newly diagnosed patients with RA and subjects in a control group, a

recent study found no signs of atherosclerosis; however after 18 months

there was a significant increase in CIMT in RA patients compared to

controls [137]. Hannnawi et el [138] found that CIMT was significantly

increased in their early RA cohort compared to controls (0.64 ± 0.13 mm

versus 0.58 ± 0.09 mm, P = 0.03). They repeated their analysis after

removing patients with previous CV events and their controls and the results

had not changed (0.63 ± 0.02 mm versus 0.57 ±0.01 mm, P = 0.03). There

are few studies investigating CIMT in early RA, indicating a need for

further studies of CIMT in early and longstanding RA.

1.16 Kynurenine pathway

The kynurenine (KYN) pathway is a metabolic pathway leading to the

production of nicotinamide adenine dinucleotide (NAD) from the

degradation of the essential amino acid enzyme tryptophan (TRP) (Figure

3). The oxidation of TRP, initiating the KYN pathway, may be catalysed by

one of three enzymes – tryptophan 2,3 dioxygenase (TDO) residing

predominantly in the liver, indoleamine 2,3 dioxygenase 1 (IDO1) or a

newly-discovered, related enzyme, indoleamine 2,3 dioxygenase 2 (IDO2)

26

[139-141]. IDO1 is the predominant extra-hepatic enzyme, found in

numerous cells, including macrophages, microglia, neurons and astrocytes

[142-145]. IDO1 is stimulated by the pro-inflammatory cytokine interferon

γ (IFN ), which is activated by T lymphocytes during an immune response

[146].

Figure 1-6: Schematic diagram of the kynurenine pathway

From : Int J Tryptophan. 2009; 2: 1-19 [147]

1.17 Kynurenine in RA

Through several pathways, KYN and its metabolites induce apoptosis of

CD4+ Th1 T-cells and generation of Tregs, thereby providing negative

feedback on T-cell responses. These effects may be helpful to maintain self-

tolerance and to limit inflammatory disorders, but little is known about the

relationship between disease activity and KYN levels in RA patients.

Furthermore, increased concentrations of KYN have been linked to tumour

cell immune escape and to poor prognosis in cancer patients [148]. One may

speculate that immune down-regulation by KYN could increase the risk of

infections, especially viral infections, where Th1 T-cell responses contribute

an important part of host defence.

27

Increased IDO activity may be measured as the KYN/TRP ratio. In patients

undergoing coronary angiography for suspected coronary artery disease, an

increased KYN/TRP ratio in urine was strongly associated with major

coronary events, acute myocardial infarctions, ischemic stroke, and all-

cause mortality [149]. Similarly, an increased ratio was linked with more

advanced atherosclerotic changes in patients on chronic haemodialysis due

to kidney failure [150]. In these settings, KYN may be acting as a

biomarker of inflammation. Taken together, KYN may play various roles in

the interactions among inflammation, immune dysregulation, CVD risk and

possibly the risks of cancer and infections in RA patients.

1.18 Rationale for the study

There are several studies that have reported and demonstrated CVD in

established and early onset RA patients, CIMT progression in patients with

RA compared to the general population, and the relationship between IDO

and or KYN/TRP ratio in RA patients. However there are very few studies

reporting the association of CVD and risk factors such as inflammation as

an increased risk for future development of RA, or comparing CIMT

progression between RA and another chronic disease with similar CV risk,

or demonstrating the relationship in KYN and CVD, new CV events,

malignancy and death in RA patients. This thesis will address some of these

issues by examining CVD and risk factors in patients who later developed

RA, comparing CIMT progression in RA and T2D and investigating KYN

concentration levels in RA patients.

1.18.1 Aims

Aim 1. To investigate the effects of CV risk factors, and CV events on the

development of RA, using a population-based cohort study design. As a

control to assess whether any findings were specific to RA or related to

arthritis in general, a parallel investigation was undertaken in participants

with and without osteoarthritis.

Aim 2. To compare the characteristics of cardiovascular risk and

progression of carotid intima-media thickness (CIMT) in individuals with

RA and type 2 diabetes.

28

Aim 3. To measure serum kynurenine in baseline samples from an RA

cohort with 10 year follow-up for cardiovascular events, and to investigate

the relationship between kynurenine concentrations, malignancy, infection,

cardiovascular disease and cause of death.

These aims will be discussed in 3 separate chapters and will form the body

of this thesis.

29

Chapter 2: Cardiovascular disease is increased prior

to onset of rheumatoid arthritis but not

osteoarthritis: HUNT population-based Norwegian

health surveys.

Published in Arthritis Research & Therapy April 2014

Helen Pahau1, Matthew A. Brown1, Sanjoy Paul*2, Ranjeny Thomas*1 and

Vibeke Videm MD*3

1University of Queensland Diamantina Institute, Translational Research

Institute, Princess Alexandra Hospital, Woolloongabba, Queensland,

Australia, 2Queensland Clinical Trials and Biostatistics Centre, School of

Population Health, University of Queensland, Princess Alexandra

Hospital, Queensland, Australia, 3Department of Laboratory Medicine,

Children’s and Women’s Health, Norwegian University of Science and

Technology, and Department of Immunology and Transfusion Medicine,

Trondheim University Hospital, Trondheim, Norway.

Full article (PDF) in appendices

2.1 Abstract

Objective: Patients with rheumatoid arthritis (RA) have increased risk of

cardiovascular (CV) events. We sought to test the hypothesis that due to

increased inflammation, CV disease and risk factors are associated with

increased risk of future RA development.

Methods: The HUNT population-based health surveys were conducted

among the entire adult population of Nord-Trøndelag, Norway to establish

the health history of 75000 people. All inhabitants 20 years or older were

invited, and information was collected through comprehensive

questionnaires, a clinical examination, and blood samples. In a cohort

design, data from HUNT2 (1995-1997, baseline) and HUNT3 (2006-2008,

follow-up) were obtained to study participants with RA (n=786) or

osteoarthritis (n=3586) at HUNT3 alone, in comparison with individuals

without RA or osteoarthritis at both times (n = 33,567).

Results: Female gender, age, smoking, body mass index, and history of

previous CV disease were associated with self-reported incident RA

(previous CV disease: odds ratio 1.52 (95% confidence interval 1.11-2.07).

30

The findings regarding previous CV disease were confirmed in sensitivity

analyses excluding participants with psoriasis (odds ratio 1.70 (1.23-2.36))

or restricting the analysis to cases with a hospital diagnosis of RA (odds

ratio 1.90 (1.10-3.27)) or carriers of the shared epitope (odds ratio 1.76

(1.13-2.74)). History of previous CV disease was not associated with

increased risk of osteoarthritis (odds ratio 1.04 (0.86-1.27)).

Conclusion: A history of previous CV disease was associated with

increased risk of incident RA but not osteoarthritis.

2.2 Introduction

Rheumatoid Arthritis (RA) is a systemic, inflammatory autoimmune

disorder that primarily involves the joints. Despite treatment advances, RA

patients continue to have higher mortality and morbidity rates than the

general population, predominantly related to increased atherosclerotic

cardiovascular (CV) disease [77, 151]. The traditional CV risk factors

include age, gender, family history, smoking, diabetes, hypertension,

dyslipidemia, obesity and sedentary lifestyle. RA patients carry a higher

burden of some of these factors. The disease occurs most commonly after 40

years of age and is associated with accelerated cellular ageing [152].

Smoking history is associated both with RA development and severity [43,

153, 154]. Physical inactivity is common in patients with RA and may

increase co-morbidity due to obesity or diabetes, leading to an increased risk

of CV mortality [155]. Some studies have also suggested that myocardial

disease without atherosclerosis is more prominent in RA patients. For

example, an echocardiography study demonstrated that RA patients

asymptomatic for CV disease had ischemia more frequently than controls,

but they also had fewer significant angiographic coronary artery stenoses

[156]. Likewise, an autopsy study showed that RA patients had diffuse

myocardial fibrosis or non-specific myocardial degeneration more

frequently than controls after excluding individuals with other known causes

[157].

Inflammation plays a central role in the pathogenesis of atherosclerosis, and

the increase in CV disease and mortality in RA may partly be explained by

31

inflammatory factors associated with RA, even after adjustment for

traditional CV risk factors [158, 159]. We and others showed that

atherosclerosis is increased both in patients with established RA and at

presentation in patients with recent-onset RA, as determined by increased

carotid intima media thickness and plaque, and is associated with their

inflammatory burden [160, 161]. The risk for MI and possible CV death is

already increased within approximately 5 years after diagnosis of RA

depending upon age and presence of CV risk factors, resulting in a 10-year

absolute risk comparable to non-RA individuals who were 5-10 years older

[14].

It has been demonstrated that inflammation pre-dates the onset of RA [162].

These data suggest that an increased risk of CV disease might also precede

the onset of RA. The Rochester Epidemiology Project study demonstrated

that RA patients were more likely to have been hospitalized because of MI

prior to RA diagnosis [163]. However, a previous longitudinal cohort study

found no difference in rate of MI, congestive heart failure or angina between

pre-RA and control individuals [164].

We hypothesized that CV risk factors and events are more prominent in

persons with incident RA, and that this augments the risk of future RA

development by increasing inflammation. The aim of the study was

therefore to investigate the effects of CV risk factors and CV events on the

development of RA, using a population-based cohort study design. As a

control to study whether any findings were specific to RA or related to

arthritis in general, a parallel investigation was performed in participants

with and without osteoarthritis.

2.3 Patients and methods

The study participants were from the HUNT population-based health

surveys conducted in the county of Nord-Trøndelag in Norway. The county

is fairly representative for Norway as a whole, with a stable and ethnically

homogenous population (3% non-Caucasians). All inhabitants 20 years or

older were invited, and information was collected through comprehensive

32

questionnaires and a clinical examination. The HUNT2 survey has

previously been described in detail [165]. The HUNT3 survey had a similar

design. In total, about 75,000 (70% of those invited) participated in HUNT2

(1995-1997), 51,000 (54% of those invited) participated in HUNT3 (2006-

2008), and 37,071 participated in both HUNT2 and HUNT3. By design the

participants were not seen during the years between inclusion in HUNT2

and HUNT3.

The study cohort consisted of all participants in both HUNT2 and HUNT3

who answered whether they had a diagnosis of RA or not (n=36,493, i.e.

98.4% of 37,071) and this number determined the study size. Incident cases

of RA were identified, i.e. participants who reported a diagnosis of RA in

HUNT3 but not in HUNT2. We also identified the incident cases of

osteoarthritis for comparison, based on the question; “Has a doctor ever said

that you have/have had any of these diseases: degenerative joint disease

(osteoarthritis)?” The question also included the Norwegian colloquial term

for osteoarthritis. Each patient group was compared to the remaining

participants of the study cohort.

Participants of HUNT gave informed consent. Approval for the study was

obtained from the Regional Committee on Medical Research Ethics, Central

Norway, the Norwegian Data Safety Authorities, and the Norwegian

Department of Health. Ethics approval was also obtained by the Metro

South Ethics Committee Brisbane. Permission was granted from the two

primary hospitals in Nord-Trøndelag, Levanger and Namsos hospitals, and

the nearest secondary referral hospital, Trondheim University Hospital, to

link the identified RA cases in our study to the hospital diagnosis registries

for verification of diagnosis. The registered diagnoses are used for billing

and reimbursement from the national insurance scheme. The ICD-9 code

714 and ICD-10 codes M05 and M06 with sub-codes were searched for. We

did not have permission to access to the patients’ case notes in order to

check that the current criteria for RA were correctly employed.

On enrolment in HUNT2 and HUNT3 the participants completed a

questionnaire incorporating information on medical history, smoking habits,

33

and family history of CV disease, defined as a parent, sibling or child with

previous MI or stroke. Information on use of lipid-lowering medications or

non-steroid anti-inflammatory drugs was not available. Anthropometric and

clinical measures included height, weight, waist and hip circumference, and

blood pressure. Non-fasting blood samples were drawn and total cholesterol,

low-density lipoprotein (LDL) cholesterol, high density lipoprotein (HDL)

cholesterol, triglycerides and serum glucose were measured using an

autoanalyzer (Hitachi). Using DNA isolated at HUNT Biobank, participants

with self-reported RA were genotyped for the Shared Epitope [166].

Samples were genotyped using the Illumina Immunochip microarray chip,

and HLA-DRB1 genotypes then determined by imputation using the

program HLA-IMP [167, 168]. Autoantibodies (anti-citrullinated peptide

antibodies or ACPA, rheumatoid factor) and C-reactive protein were not

measured in this survey, and classification into seropositive or seronegative

RA was not possible. Hypertension was defined as systolic blood pressure ≥

140 mmHg, diastolic blood pressure ≥ 90 mmHg or use of medication.

Hypercholesterolemia was defined as total serum cholesterol > 6.2 mmol/L.

Body mass index (BMI) was calculated as weight / height2 (kg/m2). Patients

who reported using over-the-counter analgesics daily for 1 month or more

during the last year before inclusion in HUNT2 were recorded as users of

analgesics. Previous CV disease was defined as a composite of angina, MI

or stroke. The relevant questions were; “Have you had or do you have

angina pectoris?” “Have you had a stroke/brain hemorrhage?” and “Have

you had a myocardial infarction?” The question about angina also included

a Norwegian colloquial term for this diagnosis. The composite variable was

used for previous CV disease as numbers of cases were too low for analysis

of individual disease events. The level of missingness at baseline for most

key variables was < 1%, with the exception of smoking (12.8% missing

data). Missingness for smoking was evenly distributed among the subgroups

of the study cohort. Non-complete cases were omitted from analysis.

2.4 Statistical methods

Data are presented as number (percentage), mean (standard deviation), or

median (interquartile range), as appropriate. The chi-square test and the

34

Mann-Whitney U-test were used for between group comparisons of

categorical and continuous study parameters, respectively. Risk factors

associated with the development of RA and osteoarthritis between HUNT2

and HUNT3 were identified using multivariate logistic regression models.

Linearity of logits for continuous variables was checked by plotting. P-

values <0.05 were considered significant. Pearson’s correlation coefficient

R was calculated to evaluate linear correlation.

Since this was a population-based survey and RA was self-reported, we

performed several sensitivity analyses to support our main analysis. Because

previous studies have indicated that few women with RA surveyed in the

community report their diagnosis accurately [169], we repeated the analysis

in several subgroups of the incident RA cases using additional criteria to

remove false positive diagnoses, i.e. 1) excluding all who also reported

having psoriasis (n=178), 2) including only patients where the diagnostic

registries of the nearest hospitals showed a diagnosis of RA (n=216), 3)

restricting the hospital-diagnosed cases to those where the diagnosis first

occurred after 1999 or 4) after 2001, to avoid including patients who forgot

to report RA in HUNT2, or 5) including only incident RA cases who carried

the shared epitope (with and without additional exclusion due to a report of

psoriasis). To represent the CV risk factors in another way, we developed

alternative models including the Framingham risk score, which is based on

age, serum cholesterol, hypertension, smoking and diabetes [170], as well as

gender and history of CV disease. The Framingham risk score was preferred

because it assigns higher risk with diabetes and there are no age limits, and

because it may easily be calculated in large cohorts. The European

HeartScore [171] is based on risk for patients between 40 and 65 years and

requires separate input for each individual using charts or an online

calculator, which was not practical in our large cohort. To verify that the

Framingham risk score was applicable in our cohort, we calculated the PC-

based HeartScore for 50 randomly selected incident RA cases and 50

controls [172]. Parallel Framingham risk scores and HeartScores were used

to calculate an equation permitting estimation of the HeartScore in the entire

cohort from their Framingham risk scores. An alternative logistic model was

developed substituting the Framingham risk score with these estimated

35

HeartScores. In addition, we compared factors associated with incident

cases of self-reported RA with incident cases of self-reported osteoarthritis.

2.5 Results

The baseline characteristics of the participants are presented in Table 1. At

baseline (1995-1997) 33,567 participants reported not having RA, and of

this group 786 (2.34%) reported RA at follow-up (2006-2008). This

corresponds to an average annual incidence of 0.21% in women and 0.16%

in men, respectively. If restricted to participants with cases identified in the

local hospital registries (n=216), the annual incidence was 0.06% in women

and 0.04% in men. As expected, incident cases of RA were older and more

of them were current or previous smokers and had hypertension at baseline.

Incident cases of RA had significantly elevated metabolic risk factors

including blood pressure, BMI, serum cholesterol and triglyceride. The

estimated median (interquartile range) Framingham risk scores at baseline

were 11 (7, 16) and 8 (4, 13) respectively in incident cases and those who

did not develop RA. Notably, a higher proportion of incident RA cases had

a history of CV disease at baseline.

Female gender, age, smoking, BMI, and previous CV disease were more

prominent in those developing RA (odds ratio 1.52 (1.11-2.07), Table 2, I -

Main model). There was minimal change in the odds ratio for previous CV

disease with inclusion of use of analgesics in this logistic regression model,

when systolic and diastolic blood pressure were included as continuous

variables instead of the categorical variable for hypertension, with

additional adjustment for total cholesterol and HDL cholesterol

concentrations, or when a variable for physical activity (low, moderate,

high) was also included (data not shown). In all the alternative multivariate

models for incident RA where the number of patients was restricted to better

characterized subgroups, previous CV disease was significant and the odds

ratios were higher than for the main model including all self-reported cases

(Table 2).

36

Table 2-1: Characteristics of individuals without rheumatoid arthritis (RA) or osteoarthritis (OA) at baseline

All Later Did not P-value Later Did not P-value

participants developed RA develop RA developed OA develop OA

without

RA/OA

n 33567 786 32781 3586 29981

Age (years) 46 (13) 51 (13) 46 (13) ** 52 (10) 45 (13) **

Sex (female) 18207 (54.2%) 488 (62.1%) 17719 (54.1%) ** 2406 (67.1%) 15801(52.7%) **

Smoking **

**

Current 8901 (26.5%) 256 (32.6%) 8645 (26.4%) 1015 (28.3%) 7886 (26.3%)

Former 7908 (23.6%)

212 (27.0%)

7696 (23.5%)

945 (26.4%)

6963 (23.2%)

Never smoker 15017 (44.7%) 281 (35.6%) 14736 (45.0%) 1425 (39.7%) 13592 (45.3%)

Hypertension 12186 (36.3%) 339 (43.1%) 11847 (36.1%) ** 1547 (43.1%) 10639 (35.5%) **

Diabetes 454 (1.4%) 19 (2.4%) 435 (1.3%) * 61 (1.7%) 393 (1.3%)

Previous CV

disease

1060 (3.2%)

52 (6.6%)

1008 (3.1%)

**

158 (4.4%)

902 (3.0%)

**

Angina 673 (2.0%) 35 (4.5%) 638 (1.9%) ** 107 (3.0%) 566 (1.9%) **

MI 441 (1.3%) 16 (2.0%) 425(1.3%) 49 (1.4%) 392 (1.3%)

Stroke 223 (0.7%) 9 (1.1%) 214 (0.7%) 32 (0.9%) 191 (0.6%)

Family risk of

CV disease

15692 (46.7%)

414 (52.7%)

15278 (46.6%)

**

2017 (56.2%)

13675 (45.6%)

**

Framingham

risk score2

9 (4-13)

11 (7-16)

8 (4-13)

11 (7-15)

8 (3-13)

Weight (kg) 76 (14) 77 (14) 76 (14) 77 (13) 76 (14)

BMI2 (kg/m2) 25.6 (23.5-

28.1)

26.3 (24.3-

29.0)

25.6 (23.4-

28.1)

**

26.4 (24.2-29.0)

25.4 (23.3-27.8)

**

Waist

circumference

(cm)

85 (11)

86 (11)

85 (11)

*

86 (11)

85 (11)

**

Hip circumference

(cm) 102 (8) 103 (8) 102 (8) ** 103 (8) 101 (6) **

37

1Values are mean (standard deviation) or numbers (%), or 2median (interquartile range). Non-respondents were excluded. CVD=cardiovascular

disease. MI=myocardial infarction. BMI=body mass index. HDL=high density lipoprotein. * p-value <0.05, ** p-value <0.001, comparing individuals

who developed RA and did not develop RA, or comparing individuals who developed OA and did not develop OA. All blood samples were non-

fasting.

II – Alternative models A to F). A diagnosis of RA was confirmed in the

hospital registries for 216 participants (27%) who had self-reported a new

RA diagnosis in HUNT3. The characteristics of these participants were very

similar to those reported for the entire group of incident RA cases (data not

shown). In the 216 cases with a hospital diagnosis of RA, the diagnosis was

found 1-2 years following enrolment in HUNT2 for 14 cases (6.5%), after 3

years for 24 cases (11.1%), after 4 years for 15 cases (6.9 %), after 5 years

for 28 cases (13.0%), and after 6 years for 22 cases (10.2 %). For the

remaining 113 cases (52.3%), the times were distributed from 7 years

onwards to the inclusion in HUNT3 with numbers varying unsystematically

from 15 to 24 cases (6.9-11.1%) per year. Thus, there was no obvious

pattern regarding the number of years from HUNT2.

In a separate alternative multivariate logistic regression model containing

gender, previous CV disease, and the Framingham risk score as covariates,

the odds ratio for developing RA in subjects with previous CV disease was

1.65 (1.21-2.24)(p<0.01). This model also suggested the likelihood of

developing RA was 6% greater with each 1 unit increase in Framingham

risk score (odds ratio: 1.06, 95% CI: 1.05-1.08), p<0.001). The model did

Waist/hip ratio 0.84 (0.08) 0.84 (0.08) 0.84 (0.08) 0.83 (0.07) 0.84 (0.08) **

Systolic blood

pressure2 (mmHg)

131 (120-144)

133 (122-146)

131 (120-144)

*

133 (122-147)

131 (120-143)

**

Diastolic blood

pressure

(mmHg)

79 (11)

80 (11)

79 (11)

*

81 (11)

79 (11)

**

Serum cholesterol

(mmol/L)

5.8 (1.2)

6.0 (1.2)

5.8 (1.2)

**

6.1 (1.2)

5.8 (1.2)

**

HDL cholesterol

(mmol/L)

1.40 (0.39)

1.37 (0.39)

1.40 (0.39)

1.44 (0.40)

1.39 (0.38)

**

Serum

triglycerides2

(mmol/L)

1.40 (0.98-

2.07)

1.51 (1.10-

2.24)

1.40 (0.97-

2.06)

**

1.46 (1.02-2.12)

1.37 (0.95-2.04)

**

38

not change with adjustment for BMI or daily use of over-the-counter

analgesics. In the model where the Framingham risk score was substituted

with the estimated HeartScore, previous cardiovascular disease remained

significant (odds ratio: 1.70 (1.25-2.32), p<0.01). In the 100 participants

where the HeartScore was calculated using the PC-based calculator, the

HeartScore (logarithmically transformed) was highly correlated with the

Framingham risk score (R=0.86, p<0.001). The correlation was very similar

in patients (R=0.89) and controls (R=0.86).

Table 2-2: Effects of risk factors on incident RA and osteoarthritis, by multivariate logistic regression

I - Main model – Rheumatoid arthritis (n=739 patients and 30,829 controls)1

CI = Confidence Interval

Odds

Ratio

95% CI P-Value

Female gender 1.50 1.29-1.75 <0.01

Age2 1.03 1.02-1.04 <0.001

Current smoker 1.64 1.38-1.96 <0.001

Former smoker 1.32 1.10-1.59 <0.01

Hypertension 0.97 0.82-1.15 0.76

Diabetes 1.34 0.83-2.18 0.23

Body mass index2 1.04 1.02-1.06 <0.001

Previous CV disease

1.52

1.11-2.07

<0.01

39

Alternative models (sensitivity analyses) - Rheumatoid arthritis

Odds Ratio 95% CI P-Value

A – After exclusion of patients with self-reported psoriasis (n=573 patients and 30,829 controls)1

Previous CV disease3 1.70 1.23-2.36 0.001

B – Including patients with hospital diagnosis of RA (n=201 patients and 30,829 controls)1


C – Including patients with hospital diagnosis of RA after 1999 (n=178 patients and 30,829 controls)1

Previous CV disease3 2.26 1.26-3.99 <0.01

D – Including patients with hospital diagnosis of RA after 2001 (n=138 patients and 30,829 controls)1


E – Including patients carrying the Shared Epitope (n=313 patients and 30,829 controls)1


F – Including patients carrying the Shared Epitope and excluding patients with self-reported

Psoriasis (n=257 patients and 30,829 controls)1


II – Model for osteoarthritis (n = 3364 patients and 24,631 controls)1

Odds

Ratio

95% CI P-Value

Female gender 2.36 2.15-2.56 <0.001

Age2 1.05 1.05-1.06 <0.001

Current smoker 1.42 1.30-1.56 <0.001

Former smoker 1.23 1.12-1.35 <0.001

Diabetes 0.90 0.66-1.22 0.49

Body mass index2 1.06 1.05-1.07 <0.001

Previous CV disease 1.04 0.86-1.27 0.66

1 Cases with complete data, 2Continuous variable, Odds ratio per 1 unit change, 3Multivariate model including the same

variables as the main model for RA

40

At follow-up (2006-2008), 3,586 participants had developed osteoarthritis,

corresponding to an average annual incidence of 1.05% in women and

0.64% in men respectively. These patients were older and the proportion of

smokers, former smokers and of those with hypertension was higher than in

the group that did not develop osteoarthritis. Systolic and diastolic blood

pressure, BMI, hip and waist circumference, total cholesterol and

triglycerides were higher for incident cases of osteoarthritis, whereas HDL

cholesterol was lower in incident cases of osteoarthritis. Female gender, age

and smoking were associated with development of osteoarthritis, whereas

previous CV disease was not (Table 2. III).

2.6 Discussion

In the present large population-based health study we found that participants

who developed RA between HUNT2 and HUNT3 had more CV disease and

CV risk factors prior to the onset of RA (at HUNT2) compared to those

without incident RA. A history of CV disease was a significant risk factor

for incident RA, as well as the previously reported risk factors: female

gender, increasing age, increasing BMI and smoking [173]. The finding of a

positive association with the Framingham risk score may be related to

smoking and age being incorporated into the score. The supplementary

analysis showed that the Framingham score was equivalent to the estimated

HeartScore in our population even though the Framingham score was not

developed from European subjects. On the other hand, previous CV disease

events did not contribute to the risk of future osteoarthritis in our study,

indicating that the finding may be specific to RA. Smoking was associated

with incident osteoarthritis.

2.6.1 CV disease and incident RA

The design of our study has several strengths. Since the data on the CV risk

factors and events were collected at HUNT2, recall bias at HUNT3 was

greatly reduced compared to a design where patients report on these factors

when receiving an RA diagnosis at a hospital-based clinic. Furthermore, the

observation time between HUNT2 and HUNT3 was approximately 10

years. The importance of a long observation time is underscored by the

41

finding that the odds ratio for the association of incident RA with previous

CV disease increased when patients receiving a hospital diagnosis of RA

during the first years following HUNT2 were excluded. It also seems

biologically plausible that longer exposure to increased inflammation due to

CV disease further increases the risk of incident RA. Furthermore, the

statistical power is greatly improved in a population-based cohort study due

to the large number of controls, and problems defining a relevant control

group as seen with case-control studies of RA are avoided.

The major limitation of our study is that data were self-reported. Previous

studies suggest that self-reported CV diagnosis and risk factors such as

hypertension and cigarette smoking are reliable [174, 175]. However, the

high incidence rate of self-reported RA in our study compared for example

to recent Swedish data based on inpatient and non-primary outpatient care

(0.056% in women, 0.025% in men) [176], confirms that there probably

were many false-positive RA diagnoses. This would tend to decrease the

power to detect true positive results in our study, but not bias towards false

positive findings. When including only cases identified in the hospital

registries, the incidence was close to the Swedish data. Furthermore, the

incidence of RA increases with observation time [177]. Therefore different

data may not be directly comparable.

Epidemiological studies also carry a risk of reverse causation, which could

ensue if persons with undiagnosed RA in HUNT2 were erroneously

included as incident cases of RA, because RA in itself increases the risk of

CV disease and events [178]. It seems unlikely that reverse causation could

explain our findings, given that the association with previous CV disease

became stronger when the cases with shortest duration between HUNT2 and

a hospital diagnosis of RA were removed from the analysis. Furthermore,

we did not observe a tendency for a higher number of cases with a hospital

diagnosis to occur during the earliest years following HUNT2, which would

have been expected with substantial reverse causation.

While fatal CV events prior to onset of RA also limited the participants in

our study to survivors of those events, this again would lead to under-

42

estimation of the impact of CV events on development of RA, or to restrict

the population at risk of RA to those ageing with CV risk that was

insufficient to lead to premature mortality. Autoantibodies were not

measured and seropositive and seronegative cases of RA could not be

distinguished, which impacts on the generalizability of our results compared

to others, such as population-based cohorts where it was concluded that

ischemic heart disease was not increased prior to RA onset [164]. There

may also have been a selection bias regarding participation in HUNT, where

sicker people or those with a lower socioeconomic status are less likely to

join. These people have a higher risk both for CV disease and for RA, which

may have biased our findings.

The results from our various sensitivity analyses unanimously supported our

main analysis. The overall conclusion that a history of CV disease is

associated with the risk for incident RA therefore seems reliable. Although

our study may not be suitable to give an exact estimate of the size of this

increased risk, the main analysis giving an OR of approximately 1.5 is

probably conservative.

2.6.2 Risk factors for osteoarthritis

Our results identifying smoking as a risk factor for future osteoarthritis

contradict some previous studies, which report an inverse association

between smoking and osteoarthritis [179]. However, the longitudinal

prospective Clearwater study designed to identify risk factors for OA

development, which included 2,505 participants, did not support this inverse

association with smoking for any of four joint sites (knee, hand, foot or

spine) [180]. In another prospective study of 1,003 participants from the

general population there was no association between smoking and

radiologically confirmed OA at different sites [181]. The current study may

have shown an association with smoking because it is larger and included a

greater number of incident osteoarthritis diagnoses. Nevertheless, since our

study relies on self-report, replication in a cohort ascertained for

radiographic osteoarthritis and symptoms is needed. However in support of

the accuracy of our case classification, we noted that higher BMI increased

the risk for osteoarthritis development, as previously described [182].

43

2.6.3 Inflammation as a link between CV disease and incident RA

From a biological perspective based on the current studies, we hypothesize

that inflammation may contribute to the parallel development of

atherosclerosis and RA during the pre-clinical period in individuals exposed

to common RA and CV risk factors, just as this interaction accelerates

complication of RA by CV events after onset.

Figure 2-1: Proposed model for illustrating the relationship among RA, CV disease,

inflammation, atherosclerosis and predisposing factors

Inflammation is the most plausible link between previous CV disease or

BMI and increased risk for RA. It is well established that the pathogenesis

of CV disease, i.e. atherogenesis, includes chronic inflammation of the wall

of muscular arteries and that inflammation is increased during acute

coronary events, associated with unstable plaque and thrombosis [163].

Furthermore, obesity induces chronic inflammation in adipose tissue,

increasing pro-inflammatory cytokines including interleukin-6 and tumour-

necrosis factor [183]. In addition, a pro-inflammatory state is present before

the clinical onset of RA.

CRP levels were found to be higher in individuals with pre-clinical RA

compared to a control group [184]. Serum cytokines and chemokines were

44

also elevated preceding the onset of RA [185]. In at least some patients,

periodontal inflammation may precede the onset of RA symptoms, also

associated with ACPA [186]. Respiratory inflammation has also been

described in ACPA positive individuals without RA [187]. Autoantibodies

may also increase the risk of immune complex mediated vascular

inflammation [188]. Thus, CV disease and CV risk factors increase the

future risk of RA development in the context of systemic inflammation and

autoantibody development [160]. Patients with seropositive RA are at

greater risk of CV disease [43, 78], and it is possible that the RA genetic

background is more permissive to atherosclerosis when associated with the

“lifestyle” factors described here. In contrast to RA, in osteoarthritis

inflammation is local and low-grade rather than systemic, and is not

associated with autoantibodies.

Figure 1 proposes a model for the relationships among RA, CV disease,

inflammation and various predisposing factors, based on the results from the

present study, as well as previous studies demonstrating an increased risk of

CV disease after diagnosis of RA. Given that our findings were robust when

adjusting for known risk factors for RA, this hypothesis merits further

investigation in studies designed to clarify pathogenetic mechanisms, which

is not possible in an epidemiological study.

2.7 Conclusions

A history of previous CV disease was associated with increased risk of

incident RA but not osteoarthritis. A probable explanation is that increased

systemic inflammation may contribute to the parallel development of

atherosclerosis and RA during the pre-clinical period in individuals exposed

to common RA and CV risk factors, in a similar way that such interaction

accelerates CV disease in patients with established RA. The association of

previous CV events with the development of RA at the population level

suggests presentation with CV events, especially in middle-aged female

smokers or first-degree relatives of RA patients, should raise clinical

suspicion to capture cases of undiagnosed early RA. Because CV risk

45

factors are increased at onset of RA, active cardiovascular risk management

is important from the time of diagnosis.

Transition to chapter 3

We and others have previously reported that CVD was present in patients as

early as at diagnosis of RA and in established RA patients. There are few

studies comparing CIMT in RA patients compared to other diseases of

similar CV risk. Therefore with the information from this chapter we

decided to compare CIMT in RA patients to patients with T2D.

46

Chapter 3: T2D patients have a higher risk of

carotid-intima media thickness progression than RA

patients

3.1 Background

It is well known that patients with Rheumatoid Arthritis (RA) or type 2

diabetes (T2D) have increased risk of atherosclerosis and cardiovascular

disease (CVD). Carotid ultrasound measurement of the intima media

thickness (CIMT) is the most commonly used method to validate

progression of atherosclerosis.

Aim: To compare the characteristics of cardiovascular risk and progression

of CIMT in individuals with RA and T2D.

Methods: We measured CIMT using B-Mode ultrasonography of the

common carotid artery, in 290 individuals (78 RA patients and 212 T2D

patients). RA patients were attending routine clinic appointments and T1D

patients were assessed as part of a lifestyle intervention study. We carried

out a physical assessment, routine laboratory investigations and took a

history of CV risk profile in all participants. CIMT was measured twice in

individuals with RA and individuals with T2D, at a mean interval of 4.8 and

2.4 years respectively.

Results: The study comprised 290 individuals: 78 with RA and 212 with

T2D. The RA cohort was older and had a higher proportion of smokers and

previous CVD. The T2D cohort had higher BMI, diastolic blood pressure

and triglycerides, lower HDL cholesterol and more statin use.

At baseline, CIMT measurements were similar in the RA and T2D cohorts

(0.88 (0.19) mm vs. 0.86 (0.21) mm p=0.80) respectively. In an adjusted

linear regression model, RA was significantly associated with lower CIMT

at follow-up (p<0.0005). Despite a shorter follow-up, 91 % of the T2D

cohort had increased CIMT at follow-up compared to 54 % of the RA

cohort (p<0.0005). In a supplementary adjusted logistic regression analysis

47

where the outcome was CIMT progression compared to unchanged or

reduced CIMT, the OR for progression in T2D was 11.4 (5.2-25.0)

compared to RA. In the RA cohort, DMARD use at baseline was associated

with significantly lower CIMT values at follow-up (p=0.04).

Conclusion: T2D patients have a much higher risk of CIMT progression

than RA patients when adjusting for relevant risk factors and baseline

CIMT.

3.2 Introduction:

There is substantial evidence that cardiovascular (CV) risk in RA is

increased [189, 190]. However, whilst individuals with rheumatoid arthritis

(RA) have higher carotid intima media thickness (CIMT) [138, 191] and

develop more CV events than the general population [192, 193], it is not

clear how rapidly cardiovascular disease (CVD) progresses relative to other

diseases at increased CV risk. Patients with type 2 diabetes (T2D) also have

an increased risk of CVD and diabetes is an independent risk factor for CV

complications [194]. Diabetic patients without a prior history of myocardial

infarction (MI) have the same risk for future events as non-diabetic patients

who have had a prior MI [195]. Traditional CV risk factors and

inflammation contribute to this increased risk in patients with RA and

diabetes [196-199]. Obesity promotes insulin resistance, and obesity-related

inflammation contributes to metabolic impairment and the risk of

development of T2D [200, 201]. Several studies have shown associations

between CIMT and CVD [129-131]. CIMT values measured by ultrasound

have been shown to correlate with direct measurement of local and systemic

atherosclerotic burden in pathology studies and with clinical CV endpoints

[202, 203]. Current CVD burden B-mode ultrasound is also commonly used

to assess arterial structure and function by measuring arterial wall thickness,

arterial wall reactivity and arterial stiffness [204206].

We found recent-onset RA patients had increased CIMT compared to

healthy controls [138]. Similar increases were reported in patients with

long-standing RA [191] and in T2D [207]. A recent study comparing T2D

patients from the Hoorn study and RA patients from the Care study reported

48

that RA and T2D had similar CVD outcomes [208], however the pattern of

CVD manifestations differed with RA having more coronary artery disease

and T2D with more peripheral arterial disease [208]. Although RA and T2D

may share some CV risk factors, differences exist, including presence of

high-grade inflammation amongst those with RA [209]. Few studies have

compared CIMT progression in two chronic diseases with elevated CV risk.

With this in mind, the objective of the current study was to compare the rate

of progression of CIMT in RA and T2D patients drawn from the same

geographic location and in whom CIMT was measured using the same

instrumentation and analysis techniques, and to relate this to CVD risk

profile and CV events.

3.3 Materials and Methods

3.3.1 Participants

Participants consecutively attending an outpatient Rheumatology clinic at a

tertiary hospital in Brisbane Australia diagnosed with RA by a

rheumatologist according to the American college of Rheumatology (ACR)

1987 criteria [44], were recruited. A health assessment and CV

questionnaire was completed by each individual. Demographic information

was ascertained by a questionnaire and disease duration was determined by

the length of RA symptoms. Patients were treated according to a response-

driven step-up algorithm [210]. Most participants were treated with a

disease-modifying anti-rheumatic drug (DMARD) or a biologic agent at

baseline.

Participants with T2D were recruited from hospital community diabetes

clinics in the same geographic area as RA participants. The participants with

T2D participated in a randomised trial of exercise intervention for 4 weeks.

248 patients underwent screening for occult coronary artery disease using

exercise echocardiography, which excluded 25 patients. Of the 223 patients

only 212 had readable CIMT measurements [211].

All patients provided formal written consent for the carotid ultrasound

measurements. Ethical approval for the current study was obtained from the

49

Metro South Human Ethics Committee. The CIMT was reported previously

in some of the RA patients [138].

3.3.2 Cardiovascular risk factor measurement:

For the RA patients cigarette smoking status was obtained by questionnaire,

which included past and present smoking history (1). In T2D patients,

present smoking status was determined by reviewing patient medical notes.

Body mass index (BMI) was calculated as weight (kilograms) divided by

the square in height (metres). Fasting lipid, erythrosedimentation rate (ESR)

and C-reactive protein (CRP) measurements were collected by the hospital

pathology department on both the RA patients and T2D patients at baseline.

Fasting glucose, insulin and HbA1c measurements collected in T2D patients

were measured prior to and after the 4-week intervention period. Blood

analyses were conducted using standard procedures by the Queensland

Health Pathology Service (Brisbane, Australia), and all measurements were

taken after an overnight fast and at least 24 h after the last exercise session.

Insulin sensitivity [HOMAIR (homoeostasis model assessment of insulin

resistance) and QUICKI (quantitative insulin check index)] and β-cell

function [HOMAβ-Cell (homoeostasis model assessment of β-cell

function)] were determined using formulae derived previously [273, 274].

In the RA patients, diagnosis of diabetes was by a physician or use of

insulin or oral hypoglycaemic medications. Hypercholesterolemia and

hypertension were identified if the diagnosis was recorded in medical notes

by a physician or patients were taking lipid-lowering or anti-hypertensive

medications at the time of assessment.

3.3.3 Cardiovascular events

To ascertain CV events, a CV questionnaire was completed by the patients

for history, dates and treatments of MI, angina and stroke, as previously

described (2) and these events were verified by reviewing the medical

records. Patients with multiple events had only one event counted per

person. MI was identified if patients developed either of the following

clinical signs: a typical rise and fall in the levels of biochemical markers

50

consistent with myocardial necrosis, in addition to ischemic symptoms,

development of pathologic Q waves on the electrocardiogram (ECG) and/or

ECG changes indicative of ischemia (ST-segment elevation or depression);

or either new pathologic Q waves on serial ECGs or pathologic changes of

healed or healing MI [212]. Stroke was identified if patients had been

admitted to the hospital with evidence of ischemic occlusion on computed

tomography (CT) scanning, carotid endarterectomy, or symptoms of stroke,

evidence of a significant plaque on carotid ultrasound and neurologic

sequelae, with the exclusion of subarachnoid haemorrhage and space-

occupying lesions. For the exercise intervention study in patients with T2D,

those with serious co-morbidity or known CVD had been excluded [211].

CV events occurring after the baseline CIMT measurement were determined

in patients with RA and T2D by review of hospital charts.

3.4 Measurement of intima media thickness

This hospital uses protocols established by The American College of

Echocardiography (ASE) for measuring CIMT [213]. CIMT was measured

from ultrasound images as earlier described using carotid duplex scanning

and automated software [138]. Ultrasound images were frozen by ECG

triggering (top of R-wave) to minimize variability depending on changes in

the CIMT and lumen diameter occurring during the cardiac cycle [138, 214,

215]. Offline analysis of the images was performed on a workstation using

standard software for automated measurement of the intima-media thickness

(IMT) (QLab 4.2.1 with IMT plugin, version 1.1; Philips Ultrasound,

Bothell, WA, USA). The algorithm uses a maximum-slope edge-detection

technique, beginning in the vessel lumen and detecting the first maximum

slope of the change in signal for the near and far walls, and repeating the

analysis to identify the media–adventitia interface. The software calculates

the mean and standard deviation for the IMT thickness in a region-of-

interest box, placed by the observer perpendicular to the vessel walls, with

the measurements performed on the R-wave of the ECG [138].

Examination and analysis of the CIMT were performed by sonographers BH

and LS, both from the same institute. The intra-observer variation and the

co-efficient of variance was 0.01 ± 0.05 mm (5%) [138]. Data are given as

51

the total average CIMT over all measured vascular segments from anterior,

lateral and posterior views from both the right and left common carotid

arteries (n=6). Areas with plaque were not included and not quantified.

3.5 Statistical analysis

Analyses were performed with IBM SPSS (v.20), STATA (v.13.1) and the

rms package in the statistical environment R. Data are presented as mean

(SD) or median (inter-quartile range), or numbers (percentages). The t-test

or Mann-Whitney U-test was used to compare continuous variables and the

chi2 –test was used to compare frequencies between patients with RA and

T2DM.

Correlations between quantitative variables were analysed by Pearson’s

correlation coefficient. P-values <0.05 were considered significant.

We under took multivariable robust linear and logistic regression modelling.

The model included pre-defined variables to analyse the CIMT

measurements, as univariate screening carries a risk of overfitting.

Transformation to normality was performed if necessary, and baseline

CIMT was included as a covariate because we expected an association with

follow-up CIMT. For linear regression, model fit was checked by residual

plotting. For logistic regression, linearity of logits for continuous variables

was checked by addition of spline terms.

For comparison between the RA and T2DM groups, the main outcome was

the mean follow-up CIMT. Secondary analyses were performed using the

average yearly rate of change in CIMT as outcome, and alternatively using

progression (positive mean yearly rate of change) or no progression (zero or

negative mean yearly rate of change) as outcome in logistic regression. For

comparisons between the two patient groups, adjustments could only be

made for variables available in both groups, i.e. age, sex, blood lipids, BMI,

blood pressure, smoking, ever use of statins, and time to follow-up. Because

the RA and T2DM groups were not matched, we also performed sensitivity

analyses on a subgroup of the patients, matching each RA patient to a

T2DM patient of the same sex and age (+ 2 years). For 3 RA patients (2

52

women aged 74 and 75 years and 1 male aged 78 years), matching by age

was not possible and these cases were omitted from the sensitivity analyses.

We used STATA to calculate Framingham Risk Score (FRS) as an

alternative way to represent CV risk [170]. A sensitivity analysis was

undertaken for CIMT at follow-up using CIMT at baseline, ever use of

statins, RA/T2D and the FRS as covariates. Age and sex were not included

because they are used in the calculation of the FRS.

For analysis of medication effects on CIMT at follow-up in the RA patients,

a step-wise modelling approach was used in order to reduce the final

number of adjustment variables. This avoided overfitting due to a small

sample size. In the first step, only baseline CIMT and classical risk factors

were included, i.e. age, sex, diabetes (no/yes), smoking, hypertension, BMI

and total cholesterol/HDL cholesterol ratio. The significant variables were

included in the next step together with factors related to RA severity: RA

duration, baseline ESR and baseline CRP. Again, only significant factors

were carried on to the third step where we also included the variables of

main interest, i.e. ever use of statins (no/yes), ever use of prednisone

(no/yes), and baseline use of all or a combination of disease modifying anti-

rheumatic drugs (DMARDs) i.e. methotrexate, hydroxychloroquine and

sulphsalazine (no/yes). The coding of statins and prednisone was chosen

because changes in medication during follow-up may have influenced

CIMT at follow-up and detailed information on dosage was not available.

Baseline DMARD use was chosen because we hypothesized early treatment

to be beneficial. In alternative models, baseline DMARD use was

exchanged with ever use of DMARDs or DMARD use at follow-up.

3.6 Results

3.6.1 Sample characteristics

The studied population consisted of 290 individuals, 78 with RA and 212

with T2D who had all undergone carotid ultrasonography for measurement

of CIMT at two time points: mean 4.8 years apart in RA patients and 2.4

years apart in T2DM patients (p=0.001). Baseline characteristics of the

participants are presented in Table 1. The two analysis time points are

designated baseline and follow-up. Recruitment of patients occurred at a

53

new-onset or a follow-up clinic appointment, thus did not necessarily occur

at diagnosis. The median disease duration for the RA group at follow up

was 19.1 yrs (9.9-28) yr and for the T2D group was 6.4 (3.4-12.4) yrs (p=

0.0005) respectively.

The burden of CV risk factors varied between the two patient groups at

baseline. The RA group was older, and had higher proportions of females,

smokers, and previous CV events. The T2D group had higher mean BMI,

diastolic blood pressure and triglycerides, lower HDL cholesterol and higher

statin use (Table 1).

At follow-up, HDL cholesterol remained lower, and systolic and diastolic

blood pressure, BMI and triglyceride levels were higher in T2D than RA

patients (Table 1). Although LDL cholesterol had decreased at follow-up for

both RA and T2D patients, LDL cholesterol levels were higher in RA than

T2D patients (Table 1). These data indicate that CVD risk factors were

present in each group but the distribution of risk factors differed between

groups.

3.7 Comparison of CIMT in RA and T2D patients

The mean CIMT at baseline was 0.88 (0.19) mm in the RA group and 0.86

(0.21) mm in the T2DM group (p=0.80). The mean CIMT at follow-up was

0.92 (0.16) mm in the RA group and 1.12 (0.25) mm in the T2DM group

(p=0.001). In multivariable linear regression adjusted as previously

described, higher follow-up CIMT was significantly positively associated

with higher baseline CIMT (p<0.0005), age (p<0.05), diabetes (p<0.0005)

and ever use of statins (p<0.05) (Figure 1). Time to follow-up was not

significantly associated with CIMT at follow-up (p=0.94). The median FRS

was higher for T2DM 16 (16-17) than for RA 13 (12-14) (p<0.001). In the

sensitivity analysis including the FRS instead of separate cardiovascular risk

factors, the FRS was not significant (p=0.32) and there were only minimal

changes in the coefficients for the significant variables mentioned above.

The mean yearly rate of CIMT change was 0.003 (-0.009-0.024) mm in the

RA group and 0.088 (0.054-0.170) mm in the T2D group (p<0.0005). In the

regression model for yearly rate of CIMT change, the only significant

54

variables were diabetes (p<0.0005) and ever use of statins (p=0.01).

Notably, baseline CIMT was not significantly associated with the yearly rate

of CIMT change (p=0.52). With correction for multiple comparisons,

however, the differences between statin users and non-users within each

diagnosis group was no longer significant (Figure 2).

193 (91%) of the T2D patients had CIMT progression over the follow-up

period as opposed to 42 (54%) of the RA patients (p<0.0005). In logistic

regression, progression of CIMT was associated with diabetes (p=0.001,

odds ratio: 6.34 (2.17-18.50)) and ever use of statins (p<0.01, odds ratio:

2.96(1.41-6.22)). Time to follow-up was not significant (p=0.23). This

model was adjusted for baseline CIMT, hypertension, smoking and BMI.

Given the differences in age and gender among RA and T2D groups at

baseline and the potential for these to skew CIMT, we carried out a

sensitivity analysis, in which data from 75 pairs of RA and T2DM patients

matched for age and gender were compared. The average CIMT at baseline

was 0.89 (0.17) mm in the RA group and 0.90 (0.17) mm in the matched

T2D group (p=0.43). The average CIMT at follow-up was 0.92 (0.16) mm

in the RA group and 1.15 (0.23) mm in the matched T2D group (p<0.0005).

The average yearly rate of CIMT change was 0.003 (-0.002-0.012) mm in

the RA group and 0.088 (0.076-0.113) mm in the matched T2DM group

(p<0.0005). In the matched pairs, 72 (96%) of the T2D patients had CIMT

progression over the follow-up period as opposed to 40 (53.3%) of the RA

patients (p<0.0005). The findings using linear and logistic regression

modelling were very similar to those reported above for the complete study

cohort (data not shown).

3.8 Influence of medication on CIMT in RA patients

Characteristics of the RA patients are given in Table 2. At baseline, 7

women (13.5%) and 7 men (26.9%) (p=0.14) were not using DMARDs.

Using the described stepwise approach for the linear regression model for

follow-up CIMT in the RA patients, age, sex, smoking and baseline CIMT

were carried on from step 1 and no factors related to RA severity were

significant in step 2. In the final model from step 3, baseline use of

DMARDs (p=0.02), and the interaction between baseline CIMT and gender

55

(p=0.03) were significantly associated with follow-up CIMT (Figure 3).

Baseline use of DMARDs had a protective effect but ever use of prednisone

was not significant (p=0.44). If RA duration was added to the final model,

this was not significant (p=0.84). In alternative models, ever use of

DMARDs or use at follow-up were not significant contributors to follow-up

CIMT (p=0.28 and 0.27, respectively).

3.9 CV events

CV events were self-reported by our RA patients, and confirmed by

reviewing all medical records.

At baseline, 16.7% of RA and no T2DM patients had experienced a CV

event. At follow-up, 30 (24.0%) T2DM patients and 10 (13.0%) RA patients

had experienced a new CV event since baseline (p=0.09) (Table 2).

3.10 Discussion

CIMT measurement by B-mode ultrasound is a valid biomarker for CV risk

in RA and T2D patients. In the present study we demonstrated that the rate

of CIMT progression was significantly lower in RA than T2D patients,

despite an observation time twice as long in RA patients. At baseline, CIMT

measurements were similar in each group. Despite a shorter follow-up

period, a significantly greater percentage of T2D patients had CIMT

progression than RA patients. By logistic regression, where the outcome

was CIMT progression compared to unchanged or reduced CIMT, OR for

progression in T2D relative to RA was 11.4. We found diabetes and ever

use of statins but not baseline CIMT were significantly associated with

yearly rate of CIMT change. In the RA cohort, DMARD use at baseline was

associated with significantly lower CIMT values at follow-up.

3.10.1 Differences in CV risk factors in RA and T2D

In our study RA patients were compared to T2D patients obtained from an

exercise prevention study [211]. Our RA and T2D cohorts were non-

matched, and the CV risk factor distribution was significantly different.

However, the sensitivity analysis for CIMT at follow-up showed that the

FRS was not significant, which fits well with our models where no CVD

56

risk factors significantly impacted follow-up CIMT or CIMT progression.

Thus, regardless of how CVD risk factors other than diabetes were

modelled, they did not significantly influence CIMT progression. To

account for the age and gender differences, a sensitivity analysis was

undertaken where we compared data from 75 pairs of RA and T2DM

matched for age and gender. This did not alter the finding that CIMT

progression was higher in T2D than RA patients. It therefore seems unlikely

that the mentioned differences in risk factors, age or gender have biased our

findings.

Although no CVD risk factors significantly impacted follow-up CIMT or

CIMT progression in multivariate models, we found that CIMT progression

in T2D was associated with use of statins. Although this may seem counter-

intuitive, statins are generally prescribed early in treatment of T2D

according to local guidelines [216], and thus statin use is likely reflective of

the hyperlipidaemia risk in T2D and perceived higher CV risk in RA

patients [217]. Figure 2 illustrates that the relationship between CITM

progression and statin use is similar in RA and T2D patients.

At follow-up, 13% of RA patients had had a CV event compared to 24% of

T2D patients (p=0.09). Although this was not a statistically significant

difference, the current study may have been underpowered to detect

differences in CV events, given their low frequency.

3.10.2 Comparison of CIMT in RA and T2D patients

A higher follow-up CIMT was positively associated with higher baseline

CIMT as expected. The impact of the higher baseline on follow-up CIMT

was eliminated by alternatively modelling CIMT yearly rate of change.

A previous study found that yearly rate of change was higher in the first 6

years after RA diagnosis [218]. Our failure to find a significant effect of

disease duration may relate to a higher proportion of longstanding RA

patients in the current than the previous study. By logistic regression of all

patients in our study, progression of CIMT was associated with diabetes and

ever use of a statin. While the same previous study found that in RA patients

with disease duration of 6yrs or longer, the change in yearly CIMT of

57

patients on a statin was comparable to those of similar disease duration and

not on a statin, the statin effect in the current study is likely to arise from the

diabetes group and not from RA duration.

3.10.3 Effect of anti-rheumatic drugs on atherosclerosis progression

Almost all studies exploring CIMT in RA have reported progression in

CIMT, however some studies have found RA patients using TNF inhibitors

have a reduction or no progression in CIMT [218-221]. In the present study

we did not have sufficient numbers of patients (n=5) on a TNF inhibitor to

assess the impact of TNF inhibition on CIMT progression [218]. However

we found CIMT progression was reduced with DMARD treatment at

baseline. Previous studies have shown that the use of DMARDs can

improve CV risk in RA patients by influencing traditional risk factors of

atherosclerosis directly or through controlling inflammation [222-224].

Furthermore the DMARD hydroxchloroquine has shown cardio-protective

properties by reducing LDL and increasing HDL cholesterol levels [225].

Prednisone use in our study had no significant effect on CIMT progression,

although in other larger studies it is suggested that glucocorticoid use may

promote atherosclerosis [226, 227]. A recent study found an association

between glucocorticoid use and carotid plaque in patients with systemic

lupus erythematosus (SLE) [228]. In RA, the glucocorticoid dose is not

normally as high as in SLE and it is possible that lower doses of

glucocorticoids either have an anti-atherogenic effect, given the

inflammatory nature of atherosclerosis, or have no effect on CIMT [229].

CIMT and carotid plaques have been found to be predictors of future CV

events in RA, with a 2.5 increased risk of CV complications among

unilateral plaques and 4.3 among those with bilateral plaques [230].

3.10.4 Advantages of CIMT as a measure of CVD

In 2010, EULAR published 10 key recommendations regarding CV risk

screening and management in RA. Early detection of CV risk factors and

annual CV risk assessment using national guidelines was recommended

[231]. CV risk screening and management are based on CV risk score

calculators like FRS or the systemic coronary risk evaluation (SCORE)

58

[171]. However a recent study found that carotid ultrasonography was more

sensitive than SCORE for the detection of subclinical atherosclerosis in RA

[232].

The ASE defines CIMT as the combined thickness of the intima and medial

layers of the far interior wall of the carotid artery [213]. ASE has set

guidelines for measuring CIMT and states that carotid ultrasound imaging

should follow a scanning and reporting protocol, similar to that of the

Atherosclerosis Risk in Communities Study [233]. Of importance, both

cohorts were examined at the same centre using the same techniques in this

study, showing good reproducibility of CIMT measurements.

A recent study supports the use of CIMT in the assessment of CV risk in

patients with RA [234]. However, another study measuring internal carotid

artery (ICA) IMT found an association of CVD risk with rate of change in

ICA-IMT but not common carotid artery (CCA) IMT in RA, suggesting a

need to evaluate both carotid arteries [218]. The standard procedure of

measuring CIMT at this hospital is at the CCA because of its size,

superficial location, ease of accessibility and limited movement when

identifying subclinical vascular disease and evaluating CVD risk [213]. We

did not examine plaque in this study.

3.11 Limitations

This study was exploratory and the sample size was small, and thus further

studies are needed to determine the generalisability of our findings.

Nevertheless, a large effect of diabetes was observed despite this limitation.

Our study groups were not matched for CV risk factors. This is not possible

when comparing RA to diabetes, because diabetes is an independent risk

factor for CVD. However, RA patients are more likely to smoke, to be older

and to be female. Thus, without a very large sample it would be very

difficult to match all subjects for CV risk except RA. The advantages of our

study are that the groups were recruited from the same hospital. They are

from the same geographical area, similar financial background and have

experienced being at a specialist outpatient clinic at this hospital. All CIMT

measurement for both groups was completed by the same sonographers. We

59

were able to match 75 RA patients by age and gender to the T2DM and the

findings using linear and logistic regression modelling were very similar to

those reported above for the complete study cohort (data not shown).

3.12 Conclusion:

In conclusion, our data suggest that the risk of CIMT progression and CV

events in RA patients is lower than in T2DM patients when comparing

patients with similar baseline CIMT. A higher burden of CV risk factors in

T2D (including diabetes itself) and early use of DMARDs in RA may

contribute to risk reduction in RA. Future studies in larger cohorts, and

comparing ICA-IMT and carotid plaque in RA and T2D would be of

interest.

Table 3-1: Characteristics of individuals with Rheumatoid Arthritis (n=78) and individuals with T2DM (n=212)

Characteristics Rheumatoid T2DM (212) P-value Rheumatoid T2DM (212) P-value Arthritis Baseline Arthritis Follow Follow-up Baseline (78) up (78)

Age, (yrs) 61 (10) 55 (10) 0.001 66 (10) 58 (10) 0.001

Sex, (female) 52 (67%) 98 (46%) 0.002

Interval between CIMT measurements

4.8 (0.85)

2.4 (1.0)

0.001

Disease Duration 13.7 (4.5-23) 4.0 (1.0-10.0) 19.1 (9.9-28) 6.4 (3.4-12.4) 0.0005

Ever Smoker 53.0 (68%) 94.0 (44%) 0.001 57.0 (73%) NA

Hypertension 27.0 (35%) 100.0 (47%) 0.056 39.0 (51%) 64.0 (30%) 0.001

Hypercholesterolemia 35.0 (47%) 121.0 (58%) 0.094 27.0 (35%) NA

CV events ∞ 13 (16.7%) 0 0.001 10 (13.7%)* 30 (24.0%) 0.09

Statins (mg) 15.0 (19%) 83.0 (39%) 0.001 24.0 (31%) 71.0 (34%) 0.66

Weight (kg) 75 (17.0) 92 (19.0) 0.001 76 (17.0) 92 (19.0) 0.001

BMI (kg/m2) 27 (6.0) 32 (6.0) 0.001 28 (8.0) 32 (6.0) 0.001

Systolic blood

pressure (mmHg)

132 (19.0)

130 (19.0)

0.44

126 (13.0)

139 (18.0)

0.001

2 Inter-quartile ranges, T2DM=type 2 diabetes mellitus, NA=not available, ∞ myocardial infarction, angina, stroke

*New CV events exclusive of baseline CV events

60

Diastolic blood

pressure (mmHg)

79 (9.0)

82 (9.0)

0.039

77 (8.0)

80 (10.0)

0.016

Serum cholesterol

(mmol/L)

4.73 (1.62)

4.85 (1.03)

0.61

4.13 (1.77)

4.60 (1.00)

0.36

HDL cholesterol

(mmol/L)

1.5 (0.42)

1.35 (0.40)

0.003

1.44(0.81)

1.22 (0.40)

0.013

Serum triglycerides

(mmol/L)2

1.10 (0.90-1.50) 1.50 (1.10-2.20)

0.001

1.0 (0.80-1.50) 1.45 (1.10-2.10)

0.001

LDL Cholesterol

(mmol/L)

3.00(1.0) 2.70 (0.90) 0.063 2.84 (0.90) 2.60 (0.90) 0.019

HbA1c2

n/a 7.20 (6.50-8.30)

n/a 7.60 (6.80-8.60)

Table 3-2: Characteristics of RA patients (n=78) at baseline and follow up and data on CV events in RA and T2DM

RA baseline (78) T2D at P-Value RA at Follow-up T2D at P-value baseline (78) follow up (212) (126)

Early RA 19.0 (24%) RA duration 14.7 (12.0) (yrs)

19.0 (12.0) 0.000

RF 43 (58%) Positive

47 (68%) 0.44

ACCP 47 (68%) Positive

48 (71%) 0.82

SE alleles 12 (43%) CRP2 7.5 (4.0-13.0) 6.0 (3.0-14.0) 0.37

ESR2 20.0 (10.0-37.0) 17.0 (12.0-26.0) 0.24

Treatment Prednisone 34 (44%) 25 (33%) 0.088

DMARDs~ 64 (82%) 65 (83%) <0.001

Biologic 3.0 (3.8%)

DMARD† 6 (7.7%) <0.001

CV Events Diabetes 7 (9.0%) 8 (10.0%) <0.001

Angina 9 (11.5%) 0 <0.001 7 (9.0%)π 10 (4.7%) 0.83

Myocardial 9 (11.5%)

Infarction

0 <0.001 3 (4.1%)σ 14 (6.6%) 0.49

Stroke 0 0 3 (3.8%) 13 (6.1%) 0.45

ACCP=anti-cyclic citrullinated protein, CRP=C-reactive protein, ESR=erythrocyte sedimentation rate, 2 Inter-quartile

ranges, ~ever use of 1 or combination of disease modifying anti-rheumatic drug, †ever use of a biologic DMARD.

T2DM=Type2 diabetes mellitus, π analysis excluded 1 pt that had angina at baseline and follow-up, σ analysis excluded

5 patients that had MI at baseline and follow up. Some patients had more than one type of CV event.

61

Figure 3-1: Relationship between mean CIMT at follow up and baseline in T2DM and

RA

• Solid black line – Relationship between mean CIMT at follow up and baseline in

T2DM

• Broken Line – Relationship between mean CIMT at follow up and baseline in RA.

• The model was adjusted for sex, ever use of statin, smoking, and baseline triglycerides,

total cholesterol/HDL cholesterol ratio, body mass index and hypertension.

• For the figure, age was set at the median, i.e. 57 years. Data were back-transformed to

the original scale.

• Thin broken lines give 95 % confidence intervals).

62

Figure 3-2: Yearly rate of change in carotid intima-media thickness in diabetes and

rheumatoid arthritis patients

1: Diabetes patients, no statin use, 2: Diabetes patients, ever statin users, 3: Rheumatoid

arthritis patients, no statin users, 4: Rheumatoid arthritis patients, ever statin users.

Differences between diabetes and RA patients: p<0.001. Differences within diabetes and

RA diagnosis group: p=0.12 and p=0.82, respectively.

63

Figure 3-3: Carotid intima-media thickness at follow-up depending on DMARD treatment.

Relationship between mean CIMT at follow-up and baseline for patients without DMARD

(disease-modifying anti- rheumatic drug) treatment at baseline (solid line) and patients with

DMARD treatment at baseline (broken line). The model was adjusted for ever use of statin,

ever use of prednisone and smoking. For the figure, age was set at the median, i.e. 62 years.

Data were back-transformed to the original scale. Thin broken lines give 95 % confidence

intervals. Overlap of confidence intervals cannot be used to assess significance in the

multivariate model.

Transition to Chapter 4

Infectious diseases (e.g. HIV), neurological disorders (e.g. AD, HD and

ALS), affective disorders (e.g. schizophrenia, depression and anxiety),

autoimmune diseases (e.g. MS and rheumatoid arthritis), peripheral

conditions (e.g. cardiovascular disease) and malignancy (e.g.

haematological neoplasia and colorectal cancer) are reported to be

associated with the up-regulation of the kynurenine pathway. Indoleamine

2,3 dioxygenase is the first enzyme along the kynurenine pathway, and is

strongly stimulated by inflammatory molecules. The final chapter we

analysed serum KYN from a cohort of RA patients in a 10 year follow-up

study.

64

Chapter 4: Kynurenine and tryptophan

concentrations and kynurenine/tryptophan ratio in

rheumatoid arthritis patients: A 10year follow-up

study.

4.1 Background

Indoleamine 2,3 dioxygenase catalyzes the rate-limiting step of tryptophan

degradation along the kynurenine pathway. The depletion of tryptophan and

generation of kynurenines play a modulatory role in the immune response.

An increase in indoleamine 2,3 dioxygenase activity determined by

tryptophan / kynurenine ratio has been shown to be associated with

infectious diseases (e.g. AIDs), neurological disorders (e.g. amyotrophic

lateral sclerosis), affective disorders (e.g. depression, anxiety), autoimmune

disease (e.g. multiple sclerosis, rheumatoid arthritis), peripheral conditions

(e.g. cardiovascular disease) and malignancy (e.g. hematological neoplasias

and colorectal cancer).

Aim

To measure serum kynurenine in baseline samples from a rheumatoid

arthritis cohort with 10 year follow-up for cardiovascular events, and to

investigate the relationship between kynurenine concentrations, malignancy,

infection, cardiovascular disease, cause of death and disease activity in RA.

Methods

Baseline sera from 129 rheumatoid arthritis patients were analysed in 2013.

These 129 patients derived from an original cohort of 231 rheumatoid

arthritis patients recruited and followed since 2003 (baseline). Medical

records of the 231 patients were reviewed in 2013 to determine

cardiovascular risk factors, new cardiovascular events, duration of

rheumatoid arthritis and rheumatological medications between 2003 and

2013. An additional 32 healthy control sera were collected in 2003.

65

Kynurenine and tryptophan concentration levels were measured in patient

and control sera by high performance liquid chromatography (HPLC).

Results

Kynurenine concentrations at baseline were higher in the RA patients than

the healthy controls. Serum kynurenine concentrations were not related to

features of rheumatoid arthritis severity, or associated with baseline and

previous cardiovascular disease, new cardiovascular disease during follow-

up, or the two combined. Furthermore there was no association between

kynurenine and malignancies or death. The only variable associated with

death was pack-years smoking.

Conclusion

Although serum kynurenine concentrations are significantly higher in RA

than in healthy controls, they were not associated with CVD, new

malignancies or death.

66

4.2 Introduction

It is now well recognized that patients with rheumatoid arthritis (RA) have

increased morbidity and mortality due to premature cardiovascular (CV)

disease (CVD). There is a 1.5-fold increase in the standardized mortality

ratio due to CV events in RA patients compared with the general population

[235]. Up to 50% of this excess CV mortality results from ischemic heart

disease (IHD), and this is closely followed by cerebrovascular disease [102,

236, 237]. Immune dysregulation and systemic inflammation are believed to

be integral to the development of accelerated atherogenesis in RA patients

[198, 238, 239].

Under normal conditions, regulatory T-cells (Treg) suppress immune

activation including self-antigen-driven inflammation, thus inhibiting

autoimmunity. Some but not all studies indicate that Treg function is

impaired in RA, possibly due to the high concentrations of the pro-

inflammatory cytokine tumour-necrosis factor (TNF) [240]. Patients with

RA have increased numbers of pro-inflammatory Th17 cells [241, 242].

Some studies have shown reduced and others increased numbers of Treg

cells in RA [243]. There is also evidence that Treg may develop into Th17

cells in autoimmune arthritis [244], also supporting the concept of immune

dysregulation in RA.

Atherosclerosis is the result of chronic inflammation in the arterial walls.

Treg have been shown to inhibit atherosclerosis by affecting lipoprotein

metabolism. Depletion of Treg in mice led to a more atherogenic lipoprotein

profile and increased atherogenesis [245]. Thus, Treg impairment in RA

patients may be one mechanism underlying accelerated atherosclerosis.

Kynurenine (KYN), a breakdown product of the amino acid tryptophan

(TRP), has emerged as a molecule with immunomodulatory effects [246].

Three different enzymes catalyse the formation of KYN from TRP:

Indoleamine 2,3-dioxygenases (IDO) 1 and 2 and TRP 2,3-dioxygenase

(TDO). IDO1 can be found in macrophages, microglia, neurons and

astrocytes [143, 144, 247]. IDO1 is upregulated by certain cytokines and

67

inflammatory molecules, particularly interferon gamma (IFN- ) [247, 248].

Indoleamine 2,3 dioxygenase 2 (IDO2) is a recently-discovered enzyme and

its encoding genes are located next to IDO1 . IDO2 possess similar structure

and enzymatic activities to IDO1, but differs in its expression pattern and

signalling, and is preferentially inhibited by D-1methyl-TRP [139, 140].

TDO resides primarily in the liver and is induced by TRP or corticosteroids

[141]. Several cytokines, including TNF, activate IDO to promote KYN

formation. Through several pathways, KYN and its metabolites induce

apoptosis of CD4+ Th1 T-cells and generation of Tregs, thereby regulating

T-cell responses. A recent study demonstrated that development of

collagen-induced arthritis was associated with increased IDO activity and

enhanced tryptophan catabolism in mice. Blocking IDO with its inhibitor, 1-

Methyl-Tryptophan, aggravated the severity of arthritis and enhanced

immune responses. These findings suggest that IDO may play an important

and novel role in the negative feedback of collagen-induced arthritis and

possibly in the pathogenesis of rheumatoid arthritis [250].

In patients undergoing coronary angiography for suspected coronary artery

disease, an increased KYN/TRP ratio in urine was strongly associated with

major coronary events, acute myocardial infarctions, ischemic stroke, and

all-cause mortality [149]. Similarly, an increased ratio was linked with more

advanced atherosclerotic changes in patients on chronic haemodialysis due

to kidney failure [150]. In these settings, KYN may be acting as a

biomarker of inflammation. Taken together, these studies suggest that KYN

may play various roles in the interactions between inflammation, immune

dysregulation, CVD risk and possibly the risks of cancer and infections in

RA patients. Here we hypothesized that increased serum KYN in RA

patients is associated with risk of CVD, cancer, serious infections and

increased severity of RA. Therefore our aim was to investigate the

relationship between kynurenine concentrations, malignancy, infection,

CVD, cause of death and disease activity in RA.

68

4.3 Materials and methods:

4.3.1 Study Population

We recruited two hundred and thirty one RA patients who met the American

College of Rheumatology (ACR) 1987 revised criteria for the classification

of RA and presented for a scheduled appointment over a 5-month period

(July to November 2003) at a tertiary hospital rheumatology clinic. This

cohort has been previously described [251]. This cohort was followed for 10

years within this hospital clinic setting. Thirty two adult healthy controls

were recruited from another study and from laboratory staff from whom

written consent was obtained.

4.3.2 Baseline data:

At baseline, CV risk factors and events were collected by completion of a

self-administered questionnaire detailing CV history, risk factors, treatment,

and details of RA, as well as laboratory data. Each patient was clinically

evaluated, with chart review to confirm history. The data on high density

lipoprotein cholesterol (HDL), low density lipoprotein cholesterol (LDL),

triglycerides, systolic and diastolic blood pressure, weight, body mass index

(BMI), smoking and previous cardiovascular disease (CVD) were recorded

and previously described [251]. The percentage risk of CV events over the

next 5 years was estimated using the 'CV disease risk calculator' derived

from the Framingham study [251]. Serum was available from 129 patients at

baseline, of whom 16 had died. No clinical data were available from the

healthy controls.

4.3.3 10-year follow-up data:

At follow-up, hospital records of all 231 participants were reviewed to

determine new cardiovascular disease events, death, malignancy and

infection information over last 10 years.

Thirty one participants died during follow-up.

4.4 Measurement of Kynurenine

Blood samples were collected at baseline and aliquots of serum were stored

at - 80⁰C until analysis. Serum concentrations of KYN and TRP were

analyzed by high performance liquid chromatography as described [4].

69

Briefly 50 µl of serum were diluted with 50 µl of normal saline and then 10

µl of 150mM sodium acetate buffer pH 4.0 was added and left at room

temperature for 2 minutes. 27 µl (1/4 volume) of 30% TCA was added and

mixed immediately with vortex. Sample mixtures were incubated on ice for

5 minutes before spinning on the pre-cooled microfuge (4℃) with max

speed (14,000 rpm). After spinning, supernatants were used for

chromatography. Standards were prepared with both low (1 µM and 0.1

µM) and high (100 µM and 10 µM) concentration of TRP and KYN. KYN

and TRP were detected with UV detector at 360 nm and 285 nm excitation

respectively. The ratio of KYN/TRP concentrations was calculated.

This study was approved by Metro South Human Research Ethics

Committee and the Australian Health & Welfare, National Death Index

registry Human Research Ethics Committee.

4.5 Statistical Analysis:

For analysis, we used SPSS and Stata (version 13.0), and the “party”

program in R. Categorical data are presented as number and percentages and

chi square tests were used to compare baseline and follow-up frequencies.

Continuous data are presented as mean and standard deviation or median

and inter-quartile range (IQR) or 95% confidence intervals as indicated. T-

tests or Mann-Whitney U-tests were used for comparisons depending on

data distribution. Multivariate robust linear regression was used to analyze

RA concentrations. Transformation to normality was performed if

necessary, and model fit was checked by residual plotting.

For analysis of dichotomous endpoints we used multivariate logistic

regression modelling, and for analysis of death, multivariate Cox regression.

P-values less than 0.05 were considered to indicate significant differences.

The number of possible explanatory variables was high compared to the

number of positive outcomes. Therefore, we first performed random forest

analysis with bootstrapping (n=2000) for feature selection. The following

variables were included in the random forest analysis: On the next step

using logistic regression or survival analysis, only the variables with an

70

association to the outcome stronger than random variation in the random

forest analysis were included [252].

Results

The original cohort consisted of 231 RA patients [251], of whom serum

KYN concentration levels were measured in 129. The follow-up

characteristics of the cohort are in Table 1. At the 10 year follow-up, this

cohort had more smokers and a higher body mass index compared to

baseline (described previously). These patients had a higher frequency of

angina, myocardial infarction, stroke, diabetes, new and old CV events.

Table 4-1: Characteristics of RA patients

n (%)

Females 80.0 (62.0%)

Age (yrs) 62.0 (13.0)

Weight 75.0 (20.0)

Body Mass Index 28.0 (7.0)

Current smoker 29.0 (23.0%)

Duration of Rheumatoid Arthritis 14.0 (13.0)

Hypertension 39.0 (30.0%)

Hyperlipidemia 33.0 (26.0%)

Previous Cardiovascular Disease

Angina 9.0 (7.0%)

Myocardial Infarction 8.0 (6.0%)

Stroke 10.0 (8.0%)

Diabetes 16.0 (12.0%)

New Cardiovascular event 21.0 (16.0%)

Old Cardiovascular event 9.0 (7.0%)

Old and New Cardiovascular event 6.0 (5.0%)

Systolic Blood Pressure 132.0 (19.0)

71

Diastolic Blood Pressure 77.0 (11.0)

Inflammatory markers

Erythrosedimentation Rate 26.0 (23.0)

C-reactive Protein 20.0 (34.0)

Cholesterol

5.1 (1,1)

Triglycerides* 1.3 (1.0-1.9)

High Density Lipoprotein 1.5 (0.5)

Low Density Lipoprotein 3.0 (0.9)

Tryptophan 57.0 (28.0)

Kynurenine 1.21 (0.58)

Numbers (%), Mean (SD), *Median (IQR).

KYN concentrations in RA patients (median 95% CI) 1.09 (0.99-1.14) µM

were higher than in healthy controls 0.91 (0.76-1.00) µM, p=0.007 (fig 1).

TRP concentrations in RA patients 48.5 (46.0-53.0) µM were also higher

than in controls 42.0 (37.0-47.0) µM, p=0.024 (fig 2). This difference was

due to a proportion of RA patients with higher KYN levels that the upper

limit for controls (Figure 1). The ratio for KYN/TRP did not differ between

RA patients and controls (p = 0.52, figure 3).

Figure 4-1: KYN concentrations in RA patients and healthy controls

72

Figure 4-2: TRP concentrations in RA patients and healthy controls

Figure 4-3: KYN/TRP ratio of RA patients and controls

73

4.7 KYN concentrations in RA patients with and without the CV and

malignancy endpoints of the study

4.7.1 A) New CV event

Of the 129 patients with baseline KYN concentrations, 21 RA patients

experienced a new CV event in the subsequent 10 years. Baseline KYN

concentrations did not predict future CV events: in patients with an event,

median KYN was 1.06 (0.83-1.32) µM and in RA patients with no new

event 1.09 (0.97-1.17) µM, p=0.91. In RA patients with previous CVD,

median KYN was 0.99 (0.85-1.26) µM and in those without previous CVD

1.1 (0.92-1.16) µM, p=0.72. Based on these findings, logistic regression

analysis was not necessary – no adjustments could change the lack of

association of KYN with previous CVD or new CV events.

Figure 4-4: KYN concentrations in RA patients with and without a new CV event between

2003 and 2013.

74

4.7.2 B) New Malignancy

In the 21 RA patients with a new malignancy at follow-up, the median

baseline KYN concentration was 1.26 (0.84-1.46) µM and in RA patients

without a new malignancy 1.06 (0.97-1.13) µM, p=0.54.

Figure 4-5: KYN concentrations in RA patients with and without a new malignancy

between 2003 and 2013.

4.7.3 C) Death

In the 16 RA patients who died during follow-up, the median KYN

concentration was 1.04 (0.881.15) µM and in the RA patients still alive 1.09

(0.97-1.23) µM, p=1.00. Two patients with a new malignancy died, and 3

patients with a new CV event died.

Figure 4-6: KYN concentrations in RA patients dying or alive between 2003 and 2013.

75

4.8 Causes of death for the RA patients with KYN data:

There were too few cases for a meaningful analysis of the relationship

between KYN and cause of death.

Table 4-2: KYN concentrations at baseline by cause of death

CVD 7 43.75

Cancer 3 18.75

Respiratory 1 6.25

Renal Disease 1 6.25

Diabetes 1 6.25

Not classified yet 3 18.75

Total 16 100

4.8.1 Other variables (than KYN) related to death

Random forest analysis was used to select possible baseline explanatory

variables predicting death of the 129 patients. Only pack-years of smoking

and average prednisone dose were selected for logistic regression; age and

sex were included as biologically important variables (Table 3). Average

prednisone dose was not significant in logistic regression (p=0.17), which

left the best model:

Table 4-3: Other variables (than KYN) related to death

Death Odds Ratio 95%CI P-value

Age 1.06 1.00-1.12 0.07

Sex 1.48 0.40-5.47 0.56

Pack-years smoking 1.03 1.00-1.05 0.02

Death Cause Frequency Percent of deaths

76

When adjusting for age and sex, pack-years of smoking significantly

predicted death within the next 10 years with an OR of 1.03 (1.00-1.05),

p=0.02, or a 3% increase in likelihood of death per pack-year. KYN

concentration was included in this model, and as expected, was not

significant (p=0.97), confirming the findings described above.

A Cox regression model was run using deaths as an outcome. The best

model was very similar to the logistic regression model, in that the only risk

factor for death was pack-years of smoking.

Table 4-4: Model with death as an outcome

Variable Hazard Ratio 95%CI P-value

Age 1.05 1.00-1.11 0.07

Sex 1.36 0.46-4.06 0.58

Pack-years smoking 1.02 1.00-1.04 0.00

KYN 0.99 0.37-2.64 0.98

77

4.9 Discussion

Although KYN has been linked to poor prognosis in cancer, renal disease

and CVD, there is little known about the relationship between disease

activity and KYN concentration levels in patients with RA. We found KYN

to be a more sensitive marker to distinguish RA patients from the healthy

controls than TRP or KYN/TRP ratio. The results of this study demonstrate

higher KYN concentrations in RA patients compared to healthy controls,

similar to results shown in previous studies [253-255]. Like Bett et al

(1962), we found KYN levels to be unrelated to RA disease duration or

disease activity [256]. In an earlier study plasma TRP concentration was

lower in RA than controls, whilst the urinary excretion of the tryptophan

metabolites, KYN, xanthurenic acid and 3-hydroxyanthralinic was increased

[257]. More recently, concentrations of TRP were measured in sera of RA

patients and found to be lower than for healthy controls, whilst KYN did not

significantly differ from controls [255].

4.9.1 The relationship between KYN and CVD

We found no significant relationship in the RA cohort between KYN and

either presence of CVD or new CV events. Previous studies have reported

KYN imbalances (an increase in the rate of formation or a relative decrease

in its rate of degradation) in CVD [258, 259]. Urine KYN was found to be a

significant predictor of major cardiovascular events, all-cause mortality and

CVD mortality, however there was no relationship between urine KYN and

stroke or non-CVD mortality [149]. KYN/TRP ratio has been suggested to

provide a more reliable measure of IDO activation than the absolute

concentrations of KYN [260]. A previous study has demonstrated that

elevated plasma KYN/TRP ratio levels are associated with increased risk of

major coronary events (MCEs) and mortality in patients with stable CAD,

participating in the Bergen Coronary Angiography Cohort (BECAC) [261].

These studies are substantially larger than our study, which may explain the

difference in our results in that our study is underpowered.

78

4.9.2 Malignancy and KYN

There were higher levels of KYN observed in patients diagnosed with colon

carcinoma, adenoma tubule villosum, or tubular adenoma than in those from

the control group in a recent study. [262]. In another study, KYN/TRYP

ratio but not KYN were increased in colorectal cancer patients (n=66)

compared to non-cancer controls (n=37) [263].

4.9.3 Previous studies of predictors of mortality in RA.

Mortality is higher in RA patients when compared to the general population,

this excess in mortality is predominantly due to coronary artery disease,

cerebrovascular atherosclerosis and other cardiovascular complications

including heart failure [264] [265, 266]. CVD associated mortality risk is

increased in both men and women who are seropositive RA [43]. CVD

associated death could be as much as 50% higher in RA patients compared

to controls, with the risk of ischemic heart disease and cerebrovascular

disease increased to a similar degree [235]. The enhanced vascular risk is

not restricted to individuals with established RA, because increased

mortality in patients who are positive for the RF and have early

inflammatory polyarthritis has been reported [43, 76].

We found in a population health survey that CV risk is increased pre onset

of RA. RF-positivity early in RA is a strong predictor of mortality. This was

demonstrated in an inception cohort of 130 RA patients, where RF-positive

RA was associated with a 6-fold increased risk of mortality compared to

patients who were RF negative at baseline [267]. In another study, RF-

positive patients excess mortality was found to be due to respiratory and

cardiovascular diseases [268]. Chronic obstructive pulmonary disease was

the most common cause of respiratory death in the RF-positive patients.

Compared to the general population the RF-positive group had a 3-fold

increased mortality rate [Standard Mortality Rate (SMR) 3.36 (95% CI 2.02,

5.24)]. SMR for ischemic heart disease (IHD) including myocardial

infarction were elevated [SMR 1.41 (95% CI 1.04, 1.88)] for RF-positive

patients, but not in those who were persistently RF-negative [268].

79

It is well documented that smoking is a risk factor for RF-positive RA [26,

269]. In Finland they found through a questionnaire that people who were

RF-positive and had arthritis, had increased mortality from all causes [270].

After adjusting for age, gender and smoking status the increased mortality

risk persisted. They also found that people who were RF-positive in the

absence of arthritis had increased cardiovascular mortality, which remained

significant after adjusting for age, gender and smoking status.

4.10 KYN concentration level and mortality

In the present study there were no differences in KYN at baseline between

the 16 patients who died during follow-up or the patients that survived and

there were too few cases for a meaningful analysis of KYN and cause of

death. In logistic regression and Cox regression analysis, after adjusting for

age and sex, pack-years of smoking was the only significant predictor of

death in our cohort. A recent study “Epidemiology and Cost-effectiveness of

Out-Of-Hospital Cardiac Arrest in Finland (FINNRESUSCI)” found KYN

correlated with intensive care unit death, that KYN was significantly higher

in patients who died compared to those that survived, and that KYN was

higher in patients with poor outcomes compared with those with good

outcomes [271]. Even though this cohort had twice the sample size (n=245)

of our own RA cohort, it is unlikely that doubling the size of the RA cohort

would have been able to show a relationship between KYN and mortality

based on the small differences observed here.

4.11 Limitations

The major limitation of this study is the small sample size and very few

endpoints. Based on our data, one would require a very large study to get

enough statistical power to demonstrate an association between KYN and

RA complications. In the healthy controls, no clinical characteristics were

available. Thus we could not compare them further to the RA cohort. There

was a high probability that the healthy controls were much younger than the

RA patients. In the present study KYN in serum has a very narrow

concentration range, therefore this test has low capacity for questions of

prediction.

80

4.12 Conclusion

Median KYN concentrations were significantly higher in RA patients than

in healthy controls, but there was a large overlap between groups. There

were no variables in our data set that could explain the differences in KYN

concentrations among the RA patients. KYN concentrations were not

related to previous or new CVD, new malignancies or death. The only

variable predicting death in our 10 year follow-up of 129 RA patients in

logistic or Cox regression was pack-years of smoking.

81

Chapter 5: Final Discussion

Most of the work presented in this thesis has already been discussed in the

individual chapters (2, 3 and 4). The main aim of this chapter is to

cohesively bring all the findings together to highlight an overall conclusion

of the thesis, the impact of the findings and how they may lead to future

research.

In this thesis, I report that CVD and CVD risk factors are significant

predictors of RA, that the increase in CIMT is similar in patients with RA

and T2D, and that KYN concentration levels are elevated in a subset of RA

patients compared to healthy controls but are not associated with risk of

death or CVD.

The most important pathophysiologic mechanism of CV risk in RA is the

relationship between systemic inflammation and atherosclerosis. There is

substantial evidence that there are influences beyond traditional risk factors

that link RA and CVD and this has led to the notion that RA patients carry

additional risk factors for CVD. I propose that inflammation and disease

activity are the fundamental links for the 3 studies.

In chapter 2 I reported an association of previous CV events with the

development of RA in individuals exposed to common RA and CV risk

factors. The increase in CVD and CV risk factors also occurs in patients

with established RA and at those at disease onset. My results concur with

previous studies showing that female gender, increasing age, increasing

BMI and smoking were risk factors for incident RA. The obvious caveat of

this study is we have relied mainly on self-reported information to detect the

presence of RA. The positive predictive value of self-reported RA has been

studied in various populations including the elderly and disabled, and was

found to be low, ranging between 21% and 34%. (3).

82

It is well documented that patients with RA and T2D develop more CV

events than the general population, as reflected in CIMT measurements.

However it is not clear how rapidly CVD progresses in RA relative to T2D,

which has a similar increased CV risk. We demonstrated (chapter 3) that

RA patients had similar baseline CIMT to T2D patients. Previous studies

have shown that DMARDs can improve CV risk in RA patients, by

influencing traditional risk factors of atherosclerosis directly or by

controlling inflammation. In our RA cohort we reported that DMARD use at

baseline was associated with significantly lower CIMT values at follow-up

than in the T2D group at follow-up. The limitations for this study are 1. The

small sample size, and 2. The cohorts were not fully matched. In a

sensitivity analysis however, we were able to match 75 pairs and the results

were similar.

Previous studies have found that the enzyme KYN correlated with intensive

care unit death, and that KYN was significantly higher in patients who died

compared to those that survived. Furthermore, KYN was higher in patients

with poor outcomes compared with those with good outcomes (4). We did

not find similar results in our cohort of RA patients. We did not undertake a

preliminary power calculation, as the study was exploratory. However, the

effect size of elevated KYN was small and we determined that we would

require a very large study to get enough statistical power to demonstrate an

association between KYN and RA complications. In the healthy controls,

no clinical characteristics were available, thus we could not compare them

further to the RA cohort. There was a high probability that the healthy

controls were much younger than the RA patients. In the present study

KYN in serum has a very narrow concentration range, therefore this test has

low capacity for questions of prediction.

Future research

Further studies addressing predictors for cardiovascular disease are needed.

When evaluating the extent of CV risk in RA, identifying markers of

inflammation and disease activity might be useful. Given that previous CV

events were associated with the development of RA, future studies should

focus on the development of effective screening methods to identify patients

83

with high cardiovascular risk and who would benefit from close monitoring

for early signs of RA requiring intervention, as well as management of CV

risk factors. There should be clear guidelines for the management of CV risk

factors in RA, similar to guidelines developed for diabetes.

From the data presented, it is unlikely that CIMT is a useful outcome

measurement for following CVD in longitudinal studies in RA. Further

outcome studies comparing CV events in RA and T2D patients would be

more useful. However these would need to be of long duration and large

size.

Taken all together, these results emphasize the complexity of the

associations between inflammation, anti-rheumatic therapies, atherosclerotic

burden and CVD in RA.

The main contributions to the field of RA and CVD research within this

thesis were

• Demonstrating that CVD precedes the onset of RA

• Comparing CIMT in T2D and RA; T2D has a higher CIMT progression

compared to RA

• Measuring KYN concentrations in RA found levels were higher for RA

than for healthy controls.

To effectively prevent cardiovascular events in patients with RA, we must

identify RA disease characteristics which signal an increased risk of CVD.

Identification of patients at high risk will support an effective individualized

strategy of targeted therapy. Our studies indicate that patients with active

RA disease qualify for intensified action in order to prevent future CV

events.

84

References

1. Scott, D.L., et al., Long-term outcome of treating rheumatoid arthritis: results after 20 years. Lancet,

1987. 1(8542): p. 1108-11.

2. Klareskog, L., et al., Genes, environment and immunity in the development of rheumatoid arthritis.

Curr Opin Immunol, 2006. 18(6): p. 650-5.

3. Minaur, N.J., et al., Outcome after 40 years with rheumatoid arthritis: a prospective study of function, disease activity, and mortality. J Rheumatol Suppl, 2004. 69: p. 3-8.

4. Goodson, N., et al., Cardiovascular admissions and mortality in an inception cohort of patients with

rheumatoid arthritis with onset in the 1980s and 1990s. Annals of the Rheumatic Diseases, 2005.

64(11): p. 1595-601.

5. Spector, T.D., Rheumatoid arthritis. Rheum Dis Clin North Am, 1990. 16(3): p. 513-37.

6. Helmick, C.G., et al., Estimates of the prevalence of arthritis and other rheumatic conditions in the

United States. Part I. Arthritis Rheum, 2008. 58(1): p. 15-25.

7. Soderlin, M.K., et al., Annual incidence of inflammatory joint diseases in a population based study in

southern Sweden. Annals of the Rheumatic Diseases, 2002. 61(10): p. 911-5.

8. Doran, M.F., et al., Trends in incidence and mortality in rheumatoid arthritis in Rochester,

Minnesota, over a forty-year period. Arthritis Rheum, 2002. 46(3): p. 625-31.

9. Aho, K., et al., Occurrence of rheumatoid arthritis in a nationwide series of twins. J Rheumatol,

1986. 13(5): p. 899-902.

10. Silman, A.J., et al., Twin concordance rates for rheumatoid arthritis: results from a nationwide

study. Br J Rheumatol, 1993. 32(10): p. 903-7.

11. Smolik, I., et al., First-degree relatives of patients with rheumatoid arthritis exhibit high prevalence of joint symptoms. J Rheumatol, 2013. 40(6): p. 818-24.

12. Kim, S.K., et al., Greater prevalence of seropositivity for anti-cyclic citrullinated peptide antibody in unaffected first-degree relatives in multicase rheumatoid arthritis-affected families. Korean J Intern

Med, 2013. 28(1): p. 45-53.

13. MacGregor, A.J., et al., Characterizing the quantitative genetic contribution to rheumatoid arthritis

using data from twins. Arthritis Rheum, 2000. 43(1): p. 30-7.

14. Seldin, M.F., et al., The genetics revolution and the assault on rheumatoid arthritis. Arthritis Rheum,

1999. 42(6): p. 1071-9.

15. Winchester, R., Reshaping Cinderella's slipper: the shared epitope hypothesis. Arthritis Res Ther,

2006. 8(3): p. 109.

16. Gregersen, P.K., J. Silver, and R.J. Winchester, Genetic Susceptibility to Rheumatoid-Arthritis and

Human-Leukocyte Antigen Class-2 Polymorphism - the Role of Shared Conformational Determinants. American Journal of Medicine, 1988. 85(6A): p. 17-19.

17. van der Helm-van Mil, A.H., et al., The HLA-DRB1 shared epitope alleles are primarily a risk factor

for anti-cyclic citrullinated peptide antibodies and are not an independent risk factor for

development of rheumatoid arthritis. Arthritis Rheum, 2006. 54(4): p. 1117-21.

18. Nielen, M.M., et al., Antibodies to citrullinated human fibrinogen (ACF) have diagnostic and

prognostic value in early arthritis. Ann Rheum Dis, 2005. 64(8): p. 1199-204.

19. Rantapaa-Dahlqvist, S., et al., Antibodies against cyclic citrullinated peptide and IgA rheumatoid

factor predict the development of rheumatoid arthritis. Arthritis Rheum, 2003. 48(10): p. 2741-9.

20. van Gaalen, F.A., et al., Autoantibodies to cyclic citrullinated peptides predict progression to

rheumatoid arthritis in patients with undifferentiated arthritis: a prospective cohort study. Arthritis

Rheum, 2004. 50(3): p. 709-15.

21. Stahl, E.A., et al., Genome-wide association study meta-analysis identifies seven new rheumatoid

arthritis risk loci. Nat Genet, 2010. 42(6): p. 508-14.

22. Bottini, N., et al., A functional variant of lymphoid tyrosine phosphatase is associated with type I

diabetes. Nat Genet, 2004. 36(4): p. 337-8.

85

23. Jawaheer, D., et al., Screening the genome for rheumatoid arthritis susceptibility genes: a replication study and combined analysis of 512 multicase families. Arthritis Rheum, 2003. 48(4): p.

906-16.

24. Kyogoku, C., et al., Genetic association of the R620W polymorphism of protein tyrosine phosphatase

PTPN22 with human SLE. Am J Hum Genet, 2004. 75(3): p. 504-7.

25. Silman, A.J., J. Newman, and A.J. MacGregor, Cigarette smoking increases the risk of rheumatoid

arthritis. Results from a nationwide study of disease-discordant twins. Arthritis and Rheumatism,

1996. 39(5): p. 732-5.

26. Stolt, P., et al., Quantification of the influence of cigarette smoking on rheumatoid arthritis: results from a population based case-control study, using incident cases. Ann Rheum Dis, 2003. 62(9): p.

835-41.

27. Wolfe, F., The effect of smoking on clinical, laboratory, and radiographic status in rheumatoid

arthritis. J Rheumatol, 2000. 27(3): p. 630-7.

28. Symmons, D.P., et al., Blood transfusion, smoking, and obesity as risk factors for the development of

rheumatoid arthritis: results from a primary care-based incident case-control study in Norfolk,

England. Arthritis Rheum, 1997. 40(11): p. 1955-61.

29. Padyukov, L., et al., A gene-environment interaction between smoking and shared epitope genes in HLA-DR provides a high risk of seropositive rheumatoid arthritis. Arthritis Rheum, 2004. 50(10): p.

3085-92.

30. Mikuls, T.R., et al., Cigarette smoking, disease severity and autoantibody expression in African

Americans with recent-onset rheumatoid arthritis. Ann Rheum Dis, 2008. 67(11): p. 1529-34.

31. Bang, S.Y., et al., Smoking increases rheumatoid arthritis susceptibility in individuals carrying the

HLA-DRB1 shared epitope, regardless of rheumatoid factor or anti-cyclic citrullinated peptide

antibody status. Arthritis Rheum, 2010. 62(2): p. 369-77.

32. Sugiyama, D., et al., Impact of smoking as a risk factor for developing rheumatoid arthritis: a metaanalysis of observational studies. Ann Rheum Dis, 2010. 69(1): p. 70-81.

33. Klareskog, L., L. Padyukov, and L. Alfredsson, Smoking as a trigger for inflammatory rheumatic diseases. Curr Opin Rheumatol, 2007. 19(1): p. 49-54.

34. Choy, E.H. and G.S. Panayi, Cytokine pathways and joint inflammation in rheumatoid arthritis. N

Engl J Med, 2001. 344(12): p. 907-16.

35. McInnes, I.B. and G. Schett, Cytokines in the pathogenesis of rheumatoid arthritis. Nat Rev Immunol, 2007. 7(6): p. 429-42.

36. Feldmann, M., F.M. Brennan, and R.N. Maini, Role of cytokines in rheumatoid arthritis. Annu Rev

Immunol, 1996. 14: p. 397-440.

37. Ainola, M., et al., Involvement of a disintegrin and a metalloproteinase 8 (ADAM8) in osteoclastogenesis and pathological bone destruction. Ann Rheum Dis, 2009. 68(3): p. 427-34.

38. Abramson, S.B. and A. Amin, Blocking the effects of IL-1 in rheumatoid arthritis protects bone and cartilage. Rheumatology (Oxford), 2002. 41(9): p. 972-80.

39. Rindfleisch, J.A. and D. Muller, Diagnosis and management of rheumatoid arthritis. Am Fam

Physician, 2005. 72(6): p. 1037-47.

40. Sherrer, Y.S., et al., The development of disability in rheumatoid arthritis. Arthritis Rheum, 1986.

29(4): p. 494-500.

41. Turesson, C., L. Jacobsson, and U. Bergstrom, Extra-articular rheumatoid arthritis: prevalence and mortality. Rheumatology (Oxford), 1999. 38(7): p. 668-74.

42. Turesson, C., et al., Extra-articular disease manifestations in rheumatoid arthritis: incidence trends

and risk factors over 46 years. Ann Rheum Dis, 2003. 62(8): p. 722-7.

43. Wallberg-Jonsson, S., et al., Extent of inflammation predicts cardiovascular disease and overall

mortality in seropositive rheumatoid arthritis. A retrospective cohort study from disease onset. J Rheumatol, 1999. 26(12): p. 2562-71.

44. Arnett, F.C., et al., The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis and Rheumatism, 1988. 31(3): p. 315-24.

86

45. Aletaha, D., et al., 2010 rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Ann Rheum Dis,

2010. 69(9): p. 1580-8.

46. Nishimura, K., et al., Meta-analysis: diagnostic accuracy of anti-cyclic citrullinated peptide antibody

and rheumatoid factor for rheumatoid arthritis. Ann Intern Med, 2007. 146(11): p. 797-808.

47. Lee, D.M. and P.H. Schur, Clinical utility of the anti-CCP assay in patients with rheumatic diseases.

Ann Rheum Dis, 2003. 62(9): p. 870-4.

48. Steiner, G. and J. Smolen, Autoantibodies in rheumatoid arthritis and their clinical significance.

Arthritis Res, 2002. 4 Suppl 2: p. S1-5.

49. Nielen, M.M., et al., Specific autoantibodies precede the symptoms of rheumatoid arthritis: a study

of serial measurements in blood donors. Arthritis Rheum, 2004. 50(2): p. 380-6.

50. Klareskog, L., et al., A new model for an etiology of rheumatoid arthritis: smoking may trigger

HLADR (shared epitope)-restricted immune reactions to autoantigens modified by citrullination. Arthritis Rheum, 2006. 54(1): p. 38-46.

51. Pedersen, M., et al., Strong combined gene-environment effects in anti-cyclic citrullinated

peptidepositive rheumatoid arthritis: a nationwide case-control study in Denmark. Arthritis Rheum,

2007. 56(5): p. 1446-53.

52. van der Heijde, D.M., et al., Influence of prognostic features on the final outcome in rheumatoid

arthritis: a review of the literature. Semin Arthritis Rheum, 1988. 17(4): p. 284-92.

53. Skinner, M. and A.S. Cohen, Amyloid P component. Methods Enzymol, 1988. 163: p. 523-36.

54. Black, S., I. Kushner, and D. Samols, C-reactive Protein. J Biol Chem, 2004. 279(47): p. 48487-90.

55. Garlanda, C., et al., Pentraxins at the crossroads between innate immunity, inflammation, matrix

deposition, and female fertility. Annu Rev Immunol, 2005. 23: p. 337-66.

56. Tillett, W.S. and T. Francis, Serological Reactions in Pneumonia with a Non-Protein Somatic Fraction

of Pneumococcus. J Exp Med, 1930. 52(4): p. 561-71.

57. Tishler, M., D. Caspi, and M. Yaron, C-reactive protein levels in patients with rheumatoid arthritis: the impact of therapy. Clin Rheumatol, 1985. 4(3): p. 321-4.

58. Hirschfield, G.M. and M.B. Pepys, C-reactive protein and cardiovascular disease: new insights from an old molecule. QJM, 2003. 96(11): p. 793-807.

59. Aguiar, F.J., et al., C-reactive protein: clinical applications and proposals for a rational use. Rev

Assoc Med Bras, 2013. 59(1): p. 85-92.

60. Prevoo, M.L., et al., Modified disease activity scores that include twenty-eight-joint counts.

Development and validation in a prospective longitudinal study of patients with rheumatoid

arthritis. Arthritis and Rheumatism, 1995. 38(1): p. 44-8.

61. Felson, D.T., et al., American College of Rheumatology. Preliminary definition of improvement in

rheumatoid arthritis. Arthritis and Rheumatism, 1995. 38(6): p. 727-35.

62. Pinals, R.S., A.T. Masi, and R.A. Larsen, Preliminary criteria for clinical remission in rheumatoid arthritis. Arthritis and Rheumatism, 1981. 24(10): p. 1308-15.

63. Dixey, J., et al., Is it possible to predict radiological damage in early rheumatoid arthritis (RA)? A

report on the occurrence, progression, and prognostic factors of radiological erosions over the first 3

years in 866 patients from the Early RA Study (ERAS). J Rheumatol Suppl, 2004. 69: p. 48-54.

64. Ostergaard, M., B. Ejbjerg, and M. Szkudlarek, Imaging in early rheumatoid arthritis: roles of

magnetic resonance imaging, ultrasonography, conventional radiography and computed tomography. Best Pract Res Clin Rheumatol, 2005. 19(1): p. 91-116.

65. Balint, P.V., et al., Ultrasound guided versus conventional joint and soft tissue fluid aspiration in

rheumatology practice: a pilot study. J Rheumatol, 2002. 29(10): p. 2209-13.

66. Catella-Lawson, F., M.P. Reilly, and G.A. FitzGerald, Cyclooxygenase inhibitors and the antiplatelet

effects of aspirin - Reply. New England Journal of Medicine, 2002. 346(20): p. 1589-1590.

67. Edwards, C.J., et al., The changing use of disease-modifying anti-rheumatic drugs in individuals with rheumatoid arthritis from the United Kingdom General Practice Research Database. Rheumatology

(Oxford), 2005. 44(11): p. 1394-8.

87

68. Bathon, J.M., et al., A comparison of etanercept and methotrexate in patients with early rheumatoid arthritis. N Engl J Med, 2000. 343(22): p. 1586-93.

69. Weinblatt, M.E., et al., Low-dose methotrexate compared with auranofin in adult rheumatoid arthritis. A thirty-six-week, double-blind trial. Arthritis and Rheumatism, 1990. 33(3): p. 330-8.

70. Saag, K.G., et al., American College of Rheumatology 2008 recommendations for the use of nonbiologic and biologic disease-modifying antirheumatic drugs in rheumatoid arthritis. Arthritis

Rheum, 2008. 59(6): p. 762-84.

71. Emery, P. and A. Gough, Why early arthritis clinics? Br J Rheumatol, 1991. 30(4): p. 241-2.

72. Quinn, M.A. and P. Emery, Are early arthritis clinics necessary? Best Pract Res Clin Rheumatol, 2005.

19(1): p. 1-17.

73. Smolen, J.S., et al., Treating rheumatoid arthritis to target: recommendations of an international task force. Ann Rheum Dis, 2010. 69(4): p. 631-7.

74. Mikuls, T.R., et al., Mortality risk associated with rheumatoid arthritis in a prospective cohort of

older women: results from the Iowa Women's Health Study. Ann Rheum Dis, 2002. 61(11): p. 994-9.

75. Singer, R.B., Mortality in rheumatoid arthritis patients treated with or without methotrexate. J Insur

Med, 2003. 35(3-4): p. 144-9.

76. Goodson, N.J., et al., Mortality in early inflammatory polyarthritis: cardiovascular mortality is

increased in seropositive patients. Arthritis Rheum, 2002. 46(8): p. 2010-9.

77. Maradit-Kremers, H., et al., Cardiovascular death in rheumatoid arthritis: a population-based study.

Arthritis Rheum, 2005. 52(3): p. 722-32.

78. Goodson, N., et al., Cardiovascular admissions and mortality in an inception cohort of patients with

rheumatoid arthritis with onset in the 1980s and 1990s. Ann Rheum Dis, 2005. 64(11): p. 1595-601.

79. Wallberg-Jonsson, S., M.L. Ohman, and S.R. Dahlqvist, Cardiovascular morbidity and mortality in

patients with seropositive rheumatoid arthritis in Northern Sweden. J Rheumatol, 1997. 24(3): p.

445-51.

80. del Rincon, I.D., et al., High incidence of cardiovascular events in a rheumatoid arthritis cohort not explained by traditional cardiac risk factors. Arthritis Rheum, 2001. 44(12): p. 2737-45.

81. Wolfe, F., et al., The mortality of rheumatoid arthritis. Arthritis Rheum, 1994. 37(4): p. 481-94.

82. Kuhn, F.E., et al., Low-Levels of High-Density-Lipoprotein Are Associated with Endothelial

Dysfunction in Humans with Angiographically Normal Coronary-Arteries. Clinical Research, 1991.

39(2): p. A147-A147.

83. Lloyd-Jones, D.M., et al., Lifetime risk of developing coronary heart disease. Lancet, 1999.

353(9147): p. 89-92.

84. Tanaka, K., et al., A nation-wide study of atherosclerosis in infants, children and young adults in

Japan. Atherosclerosis, 1988. 72(2-3): p. 143-56.

85. Getz, G.S., When is atherosclerosis not atherosclerosis? Arterioscler Thromb Vasc Biol, 2000. 20(6):

p. 1694.

86. Greenland, P., et al., 2010 ACCF/AHA guideline for assessment of cardiovascular risk in asymptomatic adults: a report of the American College of Cardiology Foundation/American Heart

Association Task Force on Practice Guidelines. Circulation, 2010. 122(25): p. e584-636.

87. Libby, P., Molecular bases of the acute coronary syndromes. Circulation, 1995. 91(11): p. 2844-50.

88. Ross, R., The Pathogenesis of Atherosclerosis - a Perspective for the 1990s. Nature, 1993. 362(6423):

p. 801-809.

89. Landmesser, U., B. Hornig, and H. Drexler, Endothelial function: a critical determinant in

atherosclerosis? Circulation, 2004. 109(21 Suppl 1): p. II27-33.

90. Groner, J.A., M. Joshi, and J.A. Bauer, Pediatric precursors of adult cardiovascular disease:

noninvasive assessment of early vascular changes in children and adolescents. Pediatrics, 2006.

118(4): p. 1683-91.

91. Strong, J.P., et al., Prevalence and extent of atherosclerosis in adolescents and young adults:

implications for prevention from the Pathobiological Determinants of Atherosclerosis in Youth

Study. Jama, 1999. 281(8): p. 727-35.

88

92. Gabriel, S.E., Cardiovascular morbidity and mortality in rheumatoid arthritis. American Journal of Medicine, 2008. 121(10 Suppl 1): p. S9-14.

93. O'Leary, D.H., et al., Carotid-artery intima and media thickness as a risk factor for myocardial

infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. N

Engl J Med, 1999. 340(1): p. 14-22.

94. Westerhof, N. and M.F. O'Rourke, Haemodynamic basis for the development of left ventricular failure in systolic hypertension and for its logical therapy. J Hypertens, 1995. 13(9): p. 943-52.

95. Covic, A., et al., Aortic pulse wave velocity and arterial wave reflections predict the extent and severity of coronary artery disease in chronic kidney disease patients. J Nephrol, 2005. 18(4): p.

38896.

96. Yusuf, 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(9438): p. 937-52.

97. Albert, C.M., et al., Prospective study of C-reactive protein, homocysteine, and plasma lipid levels as

predictors of sudden cardiac death. Circulation, 2002. 105(22): p. 2595-9.

98. Yeh, E.T., CRP as a mediator of disease. Circulation, 2004. 109(21 Suppl 1): p. II11-4.

99. Ridker, P.M., et al., Plasma concentration of C-reactive protein and risk of developing peripheral

vascular disease. Circulation, 1998. 97(5): p. 425-8.

100. Rosenson, R.S., Statins in atherosclerosis: lipid-lowering agents with antioxidant capabilities.

Atherosclerosis, 2004. 173(1): p. 1-12.

101. Rosenson, R.S., J.D. Otvos, and D.S. Freedman, Relations of lipoprotein subclass levels and lowdensity lipoprotein size to progression of coronary artery disease in the Pravastatin Limitation of

Atherosclerosis in the Coronary Arteries (PLAC-I) trial. Am J Cardiol, 2002. 90(2): p. 89-94.

102. Van Doornum S, M.G., Wicks IP. , Accelerated atherosclerosis: an extraarticular feature of

rheumatoid arthritis? . Arthritis Rheum 2002. 46.

103. Summary of the second report of the National Cholesterol Education Program (NCEP) Expert Panel

on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel

II). Jama, 1993. 269(23): p. 3015-23.

104. Iuliano, L., et al., Radiolabeled native low-density lipoprotein injected into patients with carotid stenosis accumulates in macrophages of atherosclerotic plaque : effect of vitamin E

supplementation. Circulation, 2000. 101(11): p. 1249-54.

105. Podrez, E.A., et al., Macrophage scavenger receptor CD36 is the major receptor for LDL modified by

monocyte-generated reactive nitrogen species. J Clin Invest, 2000. 105(8): p. 1095-108.

106. Tabas, I., Consequences of cellular cholesterol accumulation: basic concepts and physiological

implications. J Clin Invest, 2002. 110(7): p. 905-11.

107. Vink, H., A.A. Constantinescu, and J.A. Spaan, Oxidized lipoproteins degrade the endothelial surface

layer : implications for platelet-endothelial cell adhesion. Circulation, 2000. 101(13): p. 1500-2.

108. Tsimikas, S., et al., Temporal increases in plasma markers of oxidized low-density lipoprotein

strongly reflect the presence of acute coronary syndromes. J Am Coll Cardiol, 2003. 41(3): p. 360-70.

109. JE, B., O.C. GT, and G. SZ, Decreased HDL2 and HDL3 cholesterol, apo A-l and apo A-II, and increased

risk of myocardial infarction. Circulation, 1992. 85: p. 22.

110. Podrez, E.A., et al., Macrophage scavenger receptor CD36 is the major receptor for LDL modified by

monocyte-generated reactive nitrogen species. Journal of Clinical Investigation, 2000. 105(8): p.

1095-108.

111. Castelli, W.P., Cardiovascular disease and multifactorial risk: challenge of the 1980s. Am Heart J,

1983. 106(5 Pt 2): p. 1191-200.

112. Gotto, A.M. and E.A. Brinton, Assessing low levels of high-density lipoprotein cholesterol as a risk

factor in coronary heart disease. J Am Coll Cardiol, 2004. 43: p. 717-724.

113. Patsch, J.R., et al., Formation of high density lipoprotein2-like particles during lipolysis of very low

density lipoproteins in vitro. Proc Natl Acad Sci U S A, 1978. 75(9): p. 4519-23.

114. Gould, A.L., et al., Cholesterol reduction yields clinical benefit: impact of statin trials. Circulation,

1998. 97(10): p. 946-52.

89

115. Gould, A.L., et al., Cholesterol reduction yields clinical benefits: meta-analysis including recent trials. Clin Ther, 2007. 29(5): p. 778-94.

116. Lerman, A., et al., Statin effects beyond lipid lowering-are they clinically relevant? European Heart Journal, 2003. 24(3): p. 225-248.

117. Davignon, J., Beneficial cardiovascular pleiotropic effects of statins. Circulation, 2004. 109(23 Suppl

1): p. III39-43.

118. Furberg, C.D., et al., Pravastatin, lipids, and major coronary events. Am J Cardiol, 1994. 73(15): p.

1133-4.

119. Vaughan, C.J., A.M. Gotto, Jr., and C.T. Basson, The evolving role of statins in the management of

atherosclerosis. J Am Coll Cardiol, 2000. 35(1): p. 1-10.

120. Crisby, M., et al., Pravastatin treatment increases collagen content and decreases lipid content, inflammation, metalloproteinases, and cell death in human carotid plaques: implications for plaque

stabilization. Circulation, 2001. 103(7): p. 926-33.

121. Jialal, I., et al., Effect of hydroxymethyl glutaryl coenzyme a reductase inhibitor therapy on high

sensitive C-reactive protein levels. Circulation, 2001. 103(15): p. 1933-5.

122. Plenge, J.K., et al., Simvastatin lowers C-reactive protein within 14 days: an effect independent of

low-density lipoprotein cholesterol reduction. Circulation, 2002. 106(12): p. 1447-52.

123. McCarey, D.W., et al., Trial of Atorvastatin in Rheumatoid Arthritis (TARA): double-blind,

randomised placebo-controlled trial. Lancet, 2004. 363(9426): p. 2015-21.

124. Ridker, P.M., et al., Inflammation, pravastatin, and the risk of coronary events after myocardial

infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events (CARE)

Investigators. Circulation, 1998. 98(9): p. 839-44.

125. Frenette, P.S., Locking a leukocyte integrin with statins. N Engl J Med, 2001. 345(19): p. 1419-21.

126. Weitz-Schmidt, G., et al., Statins selectively inhibit leukocyte function antigen-1 by binding to a

novel regulatory integrin site. Nat Med, 2001. 7(6): p. 687-92.

127. Ferro, D., et al., Simvastatin inhibits the monocyte expression of proinflammatory cytokines in

patients with hypercholesterolemia. J Am Coll Cardiol, 2000. 36(2): p. 427-31.

128. Wikstrand, J. and I. Wendelhag, Methodological considerations of ultrasound investigation of

intima-media thickness and lumen diameter. J Intern Med, 1994. 236(5): p. 555-9.

129. Poli, A., et al., Ultrasonographic measurement of the common carotid artery wall thickness in

hypercholesterolemic patients. A new model for the quantitation and follow-up of preclinical atherosclerosis in living human subjects. Atherosclerosis, 1988. 70(3): p. 253-61.

130. O'Leary, D.H., et al., Distribution and correlates of sonographically detected carotid artery disease in

the Cardiovascular Health Study. The CHS Collaborative Research Group. Stroke, 1992. 23(12): p.

1752-60.

131. Mannami, T., et al., Prevalence of asymptomatic carotid atherosclerotic lesions detected by

highresolution ultrasonography and its relation to cardiovascular risk factors in the general

population of a Japanese city: the Suita study. Stroke, 1997. 28(3): p. 518-25.

132. Bots, M.L., et al., Cardiovascular determinants of carotid artery disease. The Rotterdam Elderly

Study. Hypertension, 1992. 19(6 Pt 2): p. 717-20.

133. Allan, P.L., et al., Relationship between carotid intima-media thickness and symptomatic and asymptomatic peripheral arterial disease. The Edinburgh Artery Study. Stroke, 1997. 28(2): p.

34853.

134. Blankenhorn, D.H., et al., Beneficial effects of colestipol-niacin therapy on the common carotid

artery. Two- and four-year reduction of intima-media thickness measured by ultrasound.

Circulation, 1993. 88(1): p. 20-8.

135. Furberg, C.D., et al., Effect of lovastatin on early carotid atherosclerosis and cardiovascular events.

Asymptomatic Carotid Artery Progression Study (ACAPS) Research Group. Circulation, 1994. 90(4):

p. 1679-87.

136. Poredos, P., A. Kek, and R. Verhovec, Morphological and functional changes of the arterial wall in

subjects at risk of atherosclerosis and in patients with peripheral arterial occlusive disease. Vasa,

1997. 26(4): p. 271-6.

90

137. Sodergren, A., et al., Atherosclerosis in early rheumatoid arthritis: very early endothelial activation and rapid progression of intima media thickness. Arthritis Res Ther, 2010. 12(4): p. R158.

138. Hannawi, S., et al., Atherosclerotic disease is increased in recent-onset rheumatoid arthritis: a critical role for inflammation. Arthritis Res Ther, 2007. 9(6): p. R116.

139. Metz, R., et al., Novel tryptophan catabolic enzyme IDO2 is the preferred biochemical target of the antitumor indoleamine 2,3-dioxygenase inhibitory compound D-1-methyl-tryptophan. Cancer Res,

2007. 67(15): p. 7082-7.

140. Ball, H.J., et al., Characterization of an indoleamine 2,3-dioxygenase-like protein found in humans

and mice. Gene, 2007. 396(1): p. 203-13.

141. Salter, M. and C.I. Pogson, The role of tryptophan 2,3-dioxygenase in the hormonal control of

tryptophan metabolism in isolated rat liver cells. Effects of glucocorticoids and experimental diabetes. Biochem J, 1985. 229(2): p. 499-504.

142. Guillemin, G.J., et al., Kynurenine pathway metabolism in human astrocytes: a paradox for neuronal protection. J Neurochem, 2001. 78(4): p. 842-53.

143. Guillemin, G.J., et al., Expression of the kynurenine pathway enzymes in human microglia and

macrophages. Adv Exp Med Biol, 2003. 527: p. 105-12.

144. Guillemin, G.J., et al., Characterization of the kynurenine pathway in human neurons. J Neurosci,

2007. 27(47): p. 12884-92.

145. Guillemin, G.J., et al., Expression of indoleamine 2,3-dioxygenase and production of quinolinic acid

by human microglia, astrocytes, and neurons. Glia, 2005. 49(1): p. 15-23.

146. Schroecksnadel, K., et al., Crucial role of interferon-gamma and stimulated macrophages in

cardiovascular disease. Curr Vasc Pharmacol, 2006. 4(3): p. 205-13.

147. Chen, Y. and G.J. Guillemin, Kynurenine pathway metabolites in humans: disease and healthy States. Int J Tryptophan Res, 2009. 2: p. 1-19.

148. Prendergast, G.C., Cancer: Why tumours eat tryptophan. Nature, 2011. 478(7368): p. 192-4.

149. Pedersen, E.R., et al., Urinary excretion of kynurenine and tryptophan, cardiovascular events, and

mortality after elective coronary angiography. European Heart Journal, 2013. 34(34): p. 2689-96.

150. Kato, A., et al., Relationship between an increased serum kynurenine/tryptophan ratio and

atherosclerotic parameters in hemodialysis patients. Hemodial Int, 2010. 14(4): p. 418-24.

151. Gonzalez-Juanatey, C., et al., High prevalence of subclinical atherosclerosis in psoriatic arthritis

patients without clinically evident cardiovascular disease or classic atherosclerosis risk factors. Arthritis Rheum, 2007. 57(6): p. 1074-80.

152. Schonland, S.O., et al., Premature telomeric loss in rheumatoid arthritis is genetically determined

and involves both myeloid and lymphoid cell lineages. Proc Natl Acad Sci U S A, 2003. 100(23): p.

13471-6.

153. Silman, A.J., J. Newman, and A.J. MacGregor, Cigarette smoking increases the risk of rheumatoid

arthritis. Results from a nationwide study of disease-discordant twins. Arthritis Rheum, 1996. 39(5): p. 732-5.

154. Hazes, J.M., et al., Lifestyle and the risk of rheumatoid arthritis: cigarette smoking and alcohol

consumption. Ann Rheum Dis, 1990. 49(12): p. 980-2.

155. Sokka, T., et al., Physical inactivity in patients with rheumatoid arthritis: data from twenty-one countries in a cross-sectional, international study. Arthritis Rheum, 2008. 59(1): p. 42-50.

156. Toutouzas, K., et al., Myocardial ischaemia without obstructive coronary artery disease in rheumatoid arthritis: hypothesis-generating insights from a cross-sectional study. Rheumatology

(Oxford), 2013. 52(1): p. 76-80.

157. Koivuniemi, R., et al., Cardiovascular diseases in patients with rheumatoid arthritis. Scand J

Rheumatol, 2013. 42(2): p. 131-5.

158. Liao, K.P. and D.H. Solomon, Traditional cardiovascular risk factors, inflammation and cardiovascular risk in rheumatoid arthritis. Rheumatology (Oxford), 2013. 52(1): p. 45-52.

159. Del Rincon, I., et al., Association between carotid atherosclerosis and markers of inflammation in

rheumatoid arthritis patients and healthy subjects. Arthritis Rheum, 2003. 48(7): p. 1833-40.

91

160. Hannawi, S., et al., Atherosclerotic disease is increased in recent onset rheumatoid arthritis: a critical role for inflammation. Arthritis Res Ther, 2007. 9: p. R116.

161. Gonzalez-Juanatey, C., et al., Carotid intima-media thickness predicts the development of cardiovascular events in patients with rheumatoid arthritis. Semin Arthritis Rheum, 2009. 38(5): p.

366-71.

162. Kokkonen, H., et al., Up-regulation of cytokines and chemokines predates the onset of rheumatoid

arthritis. Arthritis Rheum, 2010. 62(2): p. 383-91.

163. McEntegart, A., H. Capell, and D. Creran, Cardiovascular risk factors, including thrombotic variables,

in a population with rheumatoid arthritis. Rheumatology, 2001. 40: p. 640-644.

164. Holmqvist, M.E., et al., No increased occurrence of ischemic heart disease prior to the onset of

rheumatoid arthritis: results from two Swedish population-based rheumatoid arthritis cohorts. Arthritis Rheum, 2009. 60(10): p. 2861-9.

165. Holmen, J., et al., The Nord-Trøndelag Health Study 1995-97 (HUNT2): Objectives, contents, methods and participation. Norsk Epidemiologi, 2003. 13: p. 19-32.

166. du Montcel, S.T., et al., New classification of HLA-DRB1 alleles supports the shared epitope

hypothesis of rheumatoid arthritis susceptibility. Arthritis Rheum, 2005. 52(4): p. 1063-8.

167. Cortes, A. and M.A. Brown, Promise and pitfalls of the Immunochip. Arthritis Res Ther, 2011. 13(1):

p. 101.

168. Dilthey, A.T., et al., HLA*IMP—an integrated framework for imputing classical HLA alleles from SNP

genotypes. Bioinformatics, 2011. 27(7): p. 968-972.

169. Walitt, B.T., et al., Validation of self-report of rheumatoid arthritis and systemic lupus

erythematosus: The Women's Health Initiative. J Rheumatol, 2008. 35(5): p. 811-8.

170. D'Agostino, R.B., Sr., et al., General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation, 2008. 117(6): p. 743-53.

171. Conroy, R.M., et al., Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J, 2003. 24(11): p. 987-1003.

172. Cron, R.Q., et al., Phenotypic and functional analysis of murine CD3+,CD4-,CD8- TCR-gamma

deltaexpressing peripheral T cells. Journal of Immunology, 1989. 142: p. 3754-3762.

173. Oliver, J.E. and A.J. Silman, Risk factors for the development of rheumatoid arthritis. Scand J

Rheumatol, 2006. 35(3): p. 169-74.

174. Kehoe, R., et al., Comparing self-reported and physician-reported medical history. Am J Epidemiol,

1994. 139(8): p. 813-8.

175. Martin, L.M., et al., Validation of self-reported chronic conditions and health services in a managed

care population. Am J Prev Med, 2000. 18(3): p. 215-8.

176. Eriksson, J.K., et al., Incidence of rheumatoid arthritis in Sweden: a nationwide population-based

assessment of incidence, its determinants, and treatment penetration. Arthritis Care Res (Hoboken),

2013. 65(6): p. 870-8.

177. Wiles, N., et al., Estimating the incidence of rheumatoid arthritis: trying to hit a moving target?

Arthritis Rheum, 1999. 42(7): p. 1339-46.

178. Liu, Z., et al., Traditional Chinese medicine syndrome-related herbal prescriptions in treatment of

malignant tumors. J Tradit Chin Med, 2013. 33(1): p. 19-26.

179. Sandmark, H., et al., Osteoarthrosis of the knee in men and women in association with overweight,

smoking, and hormone therapy. Ann Rheum Dis, 1999. 58(3): p. 151-5.

180. Wilder, F.V., B.J. Hall, and J.P. Barrett, Smoking and osteoarthritis: is there an association? The

Clearwater Osteoarthritis Study. Osteoarthritis Cartilage, 2003. 11(1): p. 29-35.

181. Hart, D.J. and T.D. Spector, Cigarette smoking and risk of osteoarthritis in women in the general population: the Chingford study. Ann Rheum Dis, 1993. 52(2): p. 93-6.

182. Kulie, T., et al., Obesity and women's health: an evidence-based review. J Am Board Fam Med, 2011. 24(1): p. 75-85.

183. Tuttle, H.A., et al., Proinflammatory cytokines are increased in type 2 diabetic women with

cardiovascular disease. J Diabetes Complications, 2004. 18(6): p. 343-51.

92

184. Masi, A., J. Aldag, and J. Sipes, Do elevated levels of serum C-reactive protein predict rheumatoid arthritis in men: correlations with pre-RA status and baseline positive rheumatoid factors. J

Rheumatol, 2001. 28: p. 2359-61.

185. Hughes-Austin, J.M., et al., Multiple cytokines and chemokines are associated with rheumatoid

arthritis-related autoimmunity in first-degree relatives without rheumatoid arthritis: Studies of the

Aetiology of Rheumatoid Arthritis (SERA). Ann Rheum Dis, 2012.

186. Chen, H.-H., et al., Association between a history of periodontitis and the risk of rheumatoid

arthritis: a nationwide, population-based, case–control study. Annals of the Rheumatic Diseases,

2012.

187. Fischer, A., et al., Lung disease with anti-CCP antibodies but not rheumatoid arthritis or connective

tissue disease. Respir Med, 2012. 106(7): p. 1040-7.

188. Kerekes, G., et al., Endothelial dysfunction and atherosclerosis in rheumatoid arthritis: a

multiparametric analysis using imaging techniques and laboratory markers of inflammation and

autoimmunity. J Rheumatol, 2008. 35(3): p. 398-406.

189. Gonzalez-Gay, M.A., C. Gonzalez-Juanatey, and J. Martin, Rheumatoid arthritis: a disease associated

with accelerated atherogenesis. Semin Arthritis Rheum, 2005. 35(1): p. 8-17.

190. Kitas, G.D. and S.E. Gabriel, Cardiovascular disease in rheumatoid arthritis: state of the art and

future perspectives. Annals of the Rheumatic Diseases, 2011. 70(1): p. 8-14.

191. Del Rincon, I., et al., Acceleration of atherosclerosis during the course of rheumatoid arthritis.

Atherosclerosis, 2007. 195(2): p. 354-60.

192. Symmons, D.P., et al., Longterm mortality outcome in patients with rheumatoid arthritis: early

presenters continue to do well. J Rheumatol, 1998. 25(6): p. 1072-7.

193. Watson, D.J., T. Rhodes, and H.A. Guess, All-cause mortality and vascular events among patients

with rheumatoid arthritis, osteoarthritis, or no arthritis in the UK General Practice Research Database. J Rheumatol, 2003. 30(6): p. 1196-202.

194. Wagenknecht, L.E., et al., Impaired glucose tolerance, type 2 diabetes, and carotid wall thickness:

the Insulin Resistance Atherosclerosis Study. Diabetes Care, 1998. 21(11): p. 1812-8.

195. Haffner, S.M., et al., 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(4): p.

229-34.

196. Haffner, S.M., Lipoprotein disorders associated with type 2 diabetes mellitus and insulin resistance.

Am J Cardiol, 2002. 90(8A): p. 55i-61i.

197. Krauss, R.M., Lipids and lipoproteins in patients with type 2 diabetes. Diabetes Care, 2004. 27(6): p.

1496-504.

198. del Rincon I, W.K., Stern MP, Freeman GL, O'Leary DH, Escalante A. , Association between carotid

atherosclerosis and markers of inflammation in rheumatoid arthritis patients and healthy subjects.

Arthritis Rheum 2003. 48: p. 1833-40.

199. Dessein, P.H., A.E. Stanwix, and B.I. Joffe, Cardiovascular risk in rheumatoid arthritis versus osteoarthritis: acute phase response related decreased insulin sensitivity and high-density

lipoprotein cholesterol as well as clustering of metabolic syndrome features in rheumatoid arthritis. Arthritis Res, 2002. 4(5): p. R5.

200. Osborn, O. and J.M. Olefsky, The cellular and signaling networks linking the immune system and

metabolism in disease. Nat Med, 2012. 18(3): p. 363-74.

201. Donath, M.Y. and S.E. Shoelson, Type 2 diabetes as an inflammatory disease. Nat Rev Immunol,

2011. 11(2): p. 98-107.

202. Pignoli, P., et al., Intimal plus medial thickness of the arterial wall: a direct measurement with

ultrasound imaging. Circulation, 1986. 74(6): p. 1399-406.

203. Lorenz, M.W., et al., Prediction of clinical cardiovascular events with carotid intima-media thickness:

a systematic review and meta-analysis. Circulation, 2007. 115(4): p. 459-67.

204. Stein, J.H., et al., Vascular age: integrating carotid intima-media thickness measurements with

global coronary risk assessment. Clinical Cardiology, 2004. 27(7): p. 388-92.

93

205. Gepner, A.D., et al., Use of carotid intima-media thickness and vascular age to modify cardiovascular risk prediction. J Am Soc Echocardiogr, 2006. 19(9): p. 1170-4.

206. Bard, R.L., et al., Effect of carotid atherosclerosis screening on risk stratification during primary cardiovascular disease prevention. Am J Cardiol, 2004. 93(8): p. 1030-2.

207. Napoli, N., et al., Increased carotid thickness in subjects with recently-diagnosed diabetes from rural Cameroon. Plos One, 2012. 7(8): p. e41316.

208. van Halm, V.P., et al., Rheumatoid arthritis versus diabetes as a risk factor for cardiovascular

disease: a cross-sectional study, the CARRE Investigation. Annals of the Rheumatic Diseases, 2009.

68(9): p. 1395-400.

209. Stamatelopoulos, K.S., et al., Atherosclerosis in rheumatoid arthritis versus diabetes: a comparative

study. Arterioscler Thromb Vasc Biol, 2009. 29(10): p. 1702-8.

210. White, D., et al., Trajectory of intensive treat-to-target disease modifying drug regimen in an observational study of an early rheumatoid arthritis cohort. BMJ Open, 2013. 3(7).

211. Hordern, M.D., et al., Determinants of changes in blood glucose response to short-term exercise

training in patients with Type 2 diabetes. Clinical Science, 2008. 115(9-10): p. 273-281.

212. Alpert, J.S., et al., Myocardial infarction redefined--a consensus document of The Joint European

Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial

infarction. J Am Coll Cardiol, 2000. 36(3): p. 959-69.

213. Stein, J.H., et al., Use of carotid ultrasound to identify subclinical vascular disease and evaluate

cardiovascular disease risk: a consensus statement from the American Society of Echocardiography Carotid Intima-Media Thickness Task Force. Endorsed by the Society for Vascular Medicine. J Am Soc

Echocardiogr, 2008. 21(2): p. 93-111; quiz 189-90.

214. Salonen, J.T. and R. Salonen, Ultrasound B-mode imaging in observational studies of atherosclerotic

progression. Circulation, 1993. 87(3 Suppl): p. II56-65.

215. Bots, M.L., et al., Isolated systolic hypertension and vessel wall thickness of the carotid artery. The

Rotterdam Elderly Study. Arteriosclerosis and Thrombosis, 1993. 13(1): p. 64-9.

216. Hatherly, K., et al., Application of Australian clinical management guidelines: the current state of

play in a sample of young people living with Type 1 diabetes in the state of New South Wales and the Australian Capital Territory. Diabetes Res Clin Pract, 2011. 93(3): p. 379-84.

217. Qi, Q., et al., Genetic predisposition to dyslipidemia and type 2 diabetes risk in two prospective

cohorts. Diabetes, 2012. 61(3): p. 745-52.

218. Giles, J.T., et al., Longitudinal predictors of progression of carotid atherosclerosis in rheumatoid

arthritis. Arthritis and Rheumatism, 2011. 63(11): p. 3216-25.

219. Del Porto, F., et al., Response to anti-tumour necrosis factor alpha blockade is associated with

reduction of carotid intima-media thickness in patients with active rheumatoid arthritis.

Rheumatology (Oxford), 2007. 46(7): p. 1111-5.

220. Gonzalez-Juanatey, C., et al., Effect of anti-tumor necrosis factor alpha therapy on the progression

of subclinical atherosclerosis in severe rheumatoid arthritis. Arthritis and Rheumatism, 2006. 55(1): p. 150-3.

221. Ferrante, A., et al., Long-term anti-tumour necrosis factor therapy reverses the progression of

carotid intima-media thickness in female patients with active rheumatoid arthritis. Rheumatology

International, 2009. 30(2): p. 193-8.

222. van Halm, V.P., et al., Disease-modifying antirheumatic drugs are associated with a reduced risk for

cardiovascular disease in patients with rheumatoid arthritis: a case control study. Arthritis Res Ther,

2006. 8(5): p. R151.

223. Choi, H.K., et al., Methotrexate and mortality in patients with rheumatoid arthritis: a prospective study. Lancet, 2002. 359(9313): p. 1173-7.

224. Westlake, S.L., et al., The effect of methotrexate on cardiovascular disease in patients with

rheumatoid arthritis: a systematic literature review. Rheumatology (Oxford), 2010. 49(2): p.

295307.

225. Munro, R., et al., Effect of disease modifying agents on the lipid profiles of patients with rheumatoid

arthritis. Ann Rheum Dis, 1997. 56(6): p. 374-7.

94

226. Raynauld, J.P., Cardiovascular mortality in rheumatoid arthritis: how harmful are corticosteroids? J Rheumatol, 1997. 24(3): p. 415-6.

227. Moreland, L.W. and J.R. O'Dell, Glucocorticoids and rheumatoid arthritis: back to the future? Arthritis and Rheumatism, 2002. 46(10): p. 2553-63.

228. Manzi, S., et al., Prevalence and risk factors of carotid plaque in women with systemic lupus

erythematosus. Arthritis and Rheumatism, 1999. 42(1): p. 51-60.

229. Ross, R., Atherosclerosis--an inflammatory disease. N Engl J Med, 1999. 340(2): p. 115-26.

230. Evans, M.R., et al., Carotid atherosclerosis predicts incident acute coronary syndromes in

rheumatoid arthritis. Arthritis Rheum, 2011. 63(5): p. 1211-20.

231. Peters, M.J. and M.T. Nurmohamed, Cardiovascular risk management in rheumatoid arthritis: are

we still waiting for the first step? Arthritis Res Ther, 2013. 15(2): p. 111.

232. Corrales, A., et al., Cardiovascular risk stratification in rheumatic diseases: carotid ultrasound is more sensitive than Coronary Artery Calcification Score to detect subclinical atherosclerosis in

patients with rheumatoid arthritis. Ann Rheum Dis, 2013. 72(11): p. 1764-70.

233. Chambless, L.E., et al., Carotid wall thickness is predictive of incident clinical stroke: the

Atherosclerosis Risk in Communities (ARIC) study. Am J Epidemiol, 2000. 151(5): p. 478-87.

234. van Sijl, A.M., et al., Carotid intima media thickness in rheumatoid arthritis as compared to control

subjects: a meta-analysis. Semin Arthritis Rheum, 2011. 40(5): p. 389-97.

235. Avina-Zubieta JA, C.H., Sadatsafavi M, Etminan M, Esdaile JM, Lacaille D. , Risk of cardiovascular

mortality in patients with rheumatoid arthritis: a meta-analysis of observational studies. Arthritis Rheum 2008. 59: p. 1690-7.

236. MJ., K., Cardiovascular disease in rheumatoid arthritis. . Curr Opin Rheumatol 2006. 18.

237. MT, N., Cardiovascular risk in rheumatoid arthritis. Autoimmun Rev 2009. 8.

238. Sattar N, M.D., Capell H, McInnes IB, Explaining how “high-grade” systemic inflammation

accelerates vascular risk in rheumatoid arthritis. Circulation Circulation 2003. 108: p. 2957-63.

239. Hannawi S, H.B., Marwick TH, Thomas R, Atherosclerotic disease is increased in recent-onset rheumatoid arthritis: a critical role for inflammation. Arthritis Res Ther, 2007. 9: p. 116.

240. Cooles, F.A., J.D. Isaacs, and A.E. Anderson, Treg cells in rheumatoid arthritis: an update. Curr

Rheumatol Rep, 2013. 15(9): p. 352.

241. Lubberts, E., Th17 cytokines and arthritis. Semin Immunopathol, 2010. 32(1): p. 43-53.

242. Peck, A. and E.D. Mellins, Breaking old paradigms: Th17 cells in autoimmune arthritis. Clin Immunol,

2009. 132(3): p. 295-304.

243. Niu, Q., et al., Disturbed Th17/Treg balance in patients with rheumatoid arthritis. Rheumatology International, 2012. 32(9): p. 2731-6.

244. Komatsu, N., et al., Pathogenic conversion of Foxp3(+) T cells into TH17 cells in autoimmune

arthritis. Nat Med, 2014. 20(1): p. 62-8.

245. Klingenberg, R., et al., Depletion of FOXP3+ regulatory T cells promotes hypercholesterolemia and

atherosclerosis. Journal of Clinical Investigation, 2013. 123(3): p. 1323-34.

246. Mandi, Y. and L. Vecsei, The kynurenine system and immunoregulation. J Neural Transm, 2012.

119(2): p. 197-209.

247. Hayaishi, O. and R. Yoshida, Specific induction of pulmonary indoleamine 2,3-dioxygenase by

bacterial lipopolysaccharide. Ciba Found Symp, 1978(65): p. 199-203.

248. Werner-Felmayer, G., et al., Characteristics of interferon induced tryptophan metabolism in human

cells in vitro. Biochim Biophys Acta, 1989. 1012(2): p. 140-7.

249. Yuasa, H.J., et al., Evolution of vertebrate indoleamine 2,3-dioxygenases. J Mol Evol, 2007. 65(6): p.

705-14.

250. Szanto, S., et al., Inhibition of indoleamine 2,3-dioxygenase-mediated tryptophan catabolism accelerates collagen-induced arthritis in mice. Arthritis Res Ther, 2007. 9(3): p. R50.

251. Carroll, L., et al., Receptor for advanced glycation end products Glycine 82 Serine polymorphism and risk of cardiovascular events in rheumatoid arthritis. Arthritis Res Ther, 2007. 9(2): p. R39.

95

252. Strobl, C., J. Malley, and G. Tutz, An introduction to recursive partitioning: rationale, application, and characteristics of classification and regression trees, bagging, and random forests. Psychol

Methods, 2009. 14(4): p. 323-48.

253. Spiera, H. and R. Vallarino, Serum kynurenine in rheumatoid arthritis. J Clin Invest, 1969. 48(5): p.

856-9.

254. Forrest, C.M., et al., Kynurenine and neopterin levels in patients with rheumatoid arthritis and osteoporosis during drug treatment. Adv Exp Med Biol, 2003. 527: p. 287-95.

255. Schroecksnadel, K., et al., Increased degradation of tryptophan in blood of patients with rheumatoid arthritis. J Rheumatol, 2003. 30(9): p. 1935-9.

256. Bett, I.M., Effect of pyridoxine on tryptophan metabolism in rheumatoid arthritis. Ann Rheum Dis,

1962. 21: p. 388-91.

257. Labadarios, D., et al., Metabolic abnormalities of tryptophan and nicotinic acid in patients with rheumatoid arthritis. Rheumatol Rehabil, 1978. 17(4): p. 227-232.

258. Myint, A.M., et al., Kynurenine pathway in major depression: evidence of impaired neuroprotection.

J Affect Disord, 2007. 98(1-2): p. 143-51.

259. Maes, M., et al., Increased neopterin and interferon-gamma secretion and lower availability of

Ltryptophan in major depression: further evidence for an immune response. Psychiatry Res, 1994.

54(2): p. 143-60.

260. Schrocksnadel, K., et al., Monitoring tryptophan metabolism in chronic immune activation. Clin

Chim Acta, 2006. 364(1-2): p. 82-90.

261. Pedersen, E.R., et al., Systemic markers of interferon-gamma-mediated immune activation and

long-term prognosis in patients with stable coronary artery disease. Arterioscler Thromb Vasc Biol,

2011. 31(3): p. 698-704.

262. Walczak, K., et al., Kynurenic acid synthesis and kynurenine aminotransferases expression in colon

derived normal and cancer cells. Scand J Gastroenterol, 2011. 46(7-8): p. 903-12.

263. Huang, A., et al., Serum tryptophan decrease correlates with immune activation and impaired

quality of life in colorectal cancer. Br J Cancer, 2002. 86(11): p. 1691-6.

264. Goodson, N. and D. Symmons, Rheumatoid arthritis in women: still associated with an increased mortality. Ann Rheum Dis, 2002. 61(11): p. 955-6.

265. Riise, T., et al., Total mortality is increased in rheumatoid arthritis. A 17-year prospective study. Clin

Rheumatol, 2001. 20(2): p. 123-7.

266. Nicola, P.J., et al., Contribution of congestive heart failure and ischemic heart disease to excess mortality in rheumatoid arthritis. Arthritis Rheum, 2006. 54(1): p. 60-7.

267. van Schaardenburg, D., et al., Outcome of rheumatoid arthritis in relation to age and rheumatoid factor at diagnosis. J Rheumatol, 1993. 20(1): p. 45-52.

268. Gonzalez, A., et al., Mortality trends in rheumatoid arthritis: the role of rheumatoid factor. J

Rheumatol, 2008. 35(6): p. 1009-14.

269. Nyhall-Wahlin, B.M., et al., Smoking is a strong risk factor for rheumatoid nodules in early

rheumatoid arthritis. Ann Rheum Dis, 2006. 65(5): p. 601-6.

270. Heliovaara, M., et al., Rheumatoid factor, chronic arthritis and mortality. Ann Rheum Dis, 1995.

54(10): p. 811-4.

271. Ristagno, G., et al., Early activation of the kynurenine pathway predicts early death and long-term

outcome in patients resuscitated from out-of-hospital cardiac arrest. J Am Heart Assoc, 2014. 3(4).

272. Oliver, J.E. and A.J. Silman, Risk factors for the development of rheumatoid arthritis. Scandinavian

Journal of Rheumatology, 2006. 35(3): p. 169-174.

273. Matthews, D.R., Hosker, J. P., Rudenski, A.S., Naylor, B.A., Treacher, D.F. and Turner,

R.C. (1985) Homeostasis model assessment:insulin resistance and β-cell function from

fasting plasma glucose and insulin concentrations in man. Diabetologia 28, 412-419

96

274. Katz, A., Nambi, S.S, Mather, K. et al. (2000) Quantitative insulin sensitivity check index: a

simple, accurate method for assessing insulin sensitivity in humans. J. Clin. Endocrinol.

Metab. 85, 2402-2410

97

Appendices

Pahau et al. Arthritis Research & Therapy 2014, 16:R85http://arthritis-research.com/content/16/2/R85

RESEARCH ARTICLE Open Access

Cardiovascular disease is increased prior to onsetof rheumatoid arthritis but not osteoarthritis: thepopulation-based Nord-Trøndelag health study(HUNT)Helen Pahau1, Matthew A Brown1, Sanjoy Paul2,5†, Ranjeny Thomas1*† and Vibeke Videm3,4†

Abstract

Introduction: Patients with rheumatoid arthritis (RA) have increased risk of cardiovascular (CV) events. We soughtto test the hypothesis that due to increased inflammation, CV disease and risk factors are associated with increasedrisk of future RA development.

Methods: The population-based Nord-Trøndelag health surveys (HUNT) were conducted among the entire adultpopulation of Nord-Trøndelag, Norway. All inhabitants 20 years or older were invited, and information was collectedthrough comprehensive questionnaires, a clinical examination, and blood samples. In a cohort design, data fromHUNT2 (1995–1997, baseline) and HUNT3 (2006–2008, follow-up) were obtained to study participants with RA(n = 786) or osteoarthritis (n = 3,586) at HUNT3 alone, in comparison with individuals without RA or osteoarthritis atboth times (n = 33,567).

Results: Female gender, age, smoking, body mass index, and history of previous CV disease were associated withself-reported incident RA (previous CV disease: odds ratio 1.52 (95% confidence interval 1.11-2.07). The findingsregarding previous CV disease were confirmed in sensitivity analyses excluding participants with psoriasis (odds ratio(OR) 1.70 (1.23-2.36)) or restricting the analysis to cases with a hospital diagnosis of RA (OR 1.90 (1.10-3.27)) orcarriers of the shared epitope (OR 1.76 (1.13-2.74)). History of previous CV disease was not associated with increasedrisk of osteoarthritis (OR 1.04 (0.86-1.27)).

Conclusion: A history of previous CV disease was associated with increased risk of incident RA but notosteoarthritis.

IntroductionRheumatoid arthritis (RA) is a systemic, inflammatoryautoimmune disorder that primarily involves the joints.Despite treatment advances, RA patients continue tohave higher mortality and morbidity rates than thegeneral population, predominantly related to increasedatherosclerotic cardiovascular (CV) disease [1,2]. Thetraditional CV risk factors include age, gender, familyhistory, smoking, diabetes, hypertension, dyslipidemia,obesity and sedentary lifestyle. RA patients carry a

* Correspondence: [email protected]†Equal contributors1University of Queensland Diamantina Institute, Translational ResearchInstitute, Princess Alexandra Hospital, Woolloongabba, QLD, AustraliaFull list of author information is available at the end of the article

© 2014 Pahau et al.; licensee BioMed Central LCommons Attribution License (http://creativecreproduction in any medium, provided the orDedication waiver (http://creativecommons.orunless otherwise stated.

higher burden of some of these factors. The disease oc-curs most commonly after 40 years of age and is associ-ated with accelerated cellular ageing [3]. A history ofsmoking is associated both with RA development andseverity [4-6]. Physical inactivity is common in patientswith RA and may increase co-morbidity due to obesityor diabetes, leading to an increased risk of CV mortality[7]. Some studies have also suggested that myocardialdisease without atherosclerosis is more prominent in RApatients. For example, an echocardiography study dem-onstrated that RA patients asymptomatic for CV diseasehad ischemia more frequently than controls, but theyalso had fewer significant angiographic coronary arterystenoses [8]. Likewise, an autopsy study showed that RApatients had diffuse myocardial fibrosis or non-specific

td. This is an Open Access article distributed under the terms of the Creativeommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andiginal work is properly credited. The Creative Commons Public Domaing/publicdomain/zero/1.0/) applies to the data made available in this article,

mailto:[email protected]

http://creativecommons.org/licenses/by/2.0

http://creativecommons.org/publicdomain/zero/1.0/

Pahau et al. Arthritis Research & Therapy 2014, 16:R85 Page 2 of 9http://arthritis-research.com/content/16/2/R85

myocardial degeneration more frequently than controlsafter excluding individuals with other known causes [9].Inflammation plays a central role in the pathogenesis

of atherosclerosis, and the increase in CV disease andmortality in RA may partly be explained by inflamma-tory factors associated with RA, even after adjustmentfor traditional CV risk factors [10,11]. We and othersshowed that atherosclerosis is increased both in patientswith established RA and at presentation in patients withrecent-onset RA, as determined by increased carotid in-tima media thickness and plaque, and is associated withtheir inflammatory burden [12,13]. The risk for myocar-dial infarction (MI) and possible CV death is already in-creased within approximately 5 years after diagnosis ofRA depending upon age and presence of CV risk factors,resulting in a 10-year absolute risk comparable to non-RA individuals who were 5 to 10 years older [14].It has been demonstrated that inflammation pre-dates

the onset of RA [15]. These data suggest that an in-creased risk of CV disease might also precede the onsetof RA. The Rochester Epidemiology Project study dem-onstrated that RA patients were more likely to have beenhospitalized because of MI prior to RA diagnosis [16].However, a previous longitudinal cohort study found nodifference in the rate of MI, congestive heart failure orangina between pre-RA and control individuals [17].We hypothesized that CV risk factors and events are

more prominent in persons with incident RA, and thatthis augments the risk of future RA development by in-creasing inflammation. The aim of the study was there-fore to investigate the effects of CV risk factors and CVevents on the development of RA, using a population-based cohort study design. As a control to study whetherany findings were specific to RA or related to arthritis ingeneral, a parallel investigation was performed in partici-pants with and without osteoarthritis.

MethodsThe study participants were from the Nord-TrøndelagHealth Study (HUNT) population-based health surveysconducted in the county of Nord-Trøndelag in Norway.The county is fairly representative for Norway as a whole,with a stable and ethnically homogenous population (3%non-Caucasians). All inhabitants 20 years or older wereinvited, and information was collected through compre-hensive questionnaires and a clinical examination. TheHUNT2 survey has previously been described in detail[18]. The HUNT3 survey had a similar design. In total,about 75,000 (70% of those invited) participated inHUNT2 (1995 to 1997), 51,000 (54% of those invited) par-ticipated in HUNT3 (2006 to 2008), and 37,071 partici-pated in both HUNT2 and HUNT3. By design, theparticipants were not seen during the years between inclu-sion in HUNT2 and HUNT3.

The study cohort consisted of all participants in bothHUNT2 and HUNT3 who answered whether they had adiagnosis of RA or not (n = 36,493, that is, 98.4% of37,071) and this number determined the study size. Inci-dent cases of RA were identified, that is, participantswho reported a diagnosis of RA in HUNT3 but not inHUNT2. We also identified the incident cases of osteo-arthritis for comparison, based on the question: “Has adoctor ever said that you have/have had any of these dis-eases: degenerative joint disease (osteoarthritis)?” Thequestion also included the Norwegian colloquial termfor osteoarthritis. Each patient group was compared tothe remaining participants of the study cohort.Participants of HUNT gave informed consent. Ap-

proval for the study was obtained from the RegionalCommittee on Medical Research Ethics, Central Norway,the Norwegian Data Safety Authorities and the NorwegianDepartment of Health. Ethics approval was also obtainedby the Metro South Ethics Committee Brisbane. Permis-sion was granted from the two primary hospitals inNord-Trøndelag, Levanger and Namsos hospitals, and thenearest secondary referral hospital, Trondheim UniversityHospital, to link the identified RA cases in our study tothe hospital diagnosis registries for verification of diag-nosis. The registered diagnoses are used for billing and re-imbursement from the national insurance scheme. Wesearched for the ICD-9 code 714 and ICD-10 codes M05and M06 with sub-codes. We did not have permission toaccess to the patients’ case notes in order to check that thecurrent criteria for RA were correctly employed.On enrolment in HUNT2 and HUNT3 the participants

completed a questionnaire incorporating information onmedical history, smoking habits and family history of CVdisease, defined as a parent, sibling or child with previousMI or stroke. Information on the use of lipid-loweringmedications or non-steroid anti-inflammatory drugs wasnot available. Anthropometric and clinical measures in-cluded height, weight, waist and hip circumference, andblood pressure. Non-fasting blood samples were drawn andtotal cholesterol, low-density lipoprotein (LDL) cholesterol,high density lipoprotein (HDL) cholesterol, triglyceridesand serum glucose were measured using an autoanalyzer(Hitachi Biocore Systems, Thornhill, ON, Canada). UsingDNA isolated at HUNT Biobank, participants with self-reported RA were genotyped for the Shared Epitope [19].Samples were genotyped using the Illumina Immunochipmicroarray chip, and HLA-DRB1 genotypes then determinedby imputation using the program HLA-IMP [20,21]. Auto-antibodies (anti-citrullinated peptide antibodies or ACPA,rheumatoid factor) and C-reactive protein were not mea-sured in this survey, and classification into seropositive orseronegative RA was not possible. Hypertension was definedas systolic blood pressure ≥140 mmHg, diastolic blood pres-sure ≥90 mmHg or use of medication. Hypercholesterolemia


was defined as total serum cholesterol >6.2 mmol/L. Bodymass index (BMI) was calculated as weight/height2 (kg/m2).Patients who reported using over-the-counter analgesicsdaily for one month or more during the last year beforeinclusion in HUNT2 were recorded as users of analgesics.Previous CV disease was defined as a composite of angina,MI or stroke. The relevant questions were; “Have you hador do you have angina pectoris?” “Have you had a stroke/brain hemorrhage?” and “Have you had a myocardialinfarction?” The question about angina also included aNorwegian colloquial term for this diagnosis. The com-posite variable was used for previous CV disease as num-bers of cases were too low for analysis of individualdisease events. The level of missingness at baseline formost key variables was <1%, with the exception of smok-ing (12.8% missing data). Missingness for smoking wasevenly distributed among the subgroups of the study co-hort. Non-complete cases were omitted from analysis.

Statistical methodsData are presented as number (percentage), mean (stand-ard deviation) or median (interquartile range), as appropri-ate. The chi-square test and the Mann-Whitney U-testwere used for between-group comparisons of categoricaland continuous study parameters, respectively. Risk factorsassociated with the development of RA and osteoarthritisbetween HUNT2 and HUNT3 were identified usingmultivariate logistic regression models. Linearity of logitsfor continuous variables was checked by plotting.P-values <0.05 were considered significant. Pearson’s cor-relation coefficient R was calculated to evaluate linearcorrelation.Since this was a population-based survey and RA was

self-reported, we performed several sensitivity analyses tosupport our main analysis. Because previous studies haveindicated that few women with RA surveyed in the com-munity report their diagnosis accurately [22], we repeatedthe analysis in several subgroups of the incident RA casesusing additional criteria to remove false-positive diag-noses, that is, 1) excluding all who also reported havingpsoriasis (n = 178), 2) including only patients where thediagnostic registries of the nearest hospitals showed adiagnosis of RA (n = 216), 3) restricting the hospital-diagnosed cases to those where the diagnosis first oc-curred after 1999 or 4) after 2001, to avoid includingpatients who forgot to report RA in HUNT2, or 5) includ-ing only incident RA cases who carried the shared epitope(with and without additional exclusion due to a report ofpsoriasis). To represent the CV risk factors in anotherway, we developed alternative models including theFramingham risk score, which is based on age, serumcholesterol, hypertension, smoking and diabetes [23], aswell as gender and history of CV disease. The Framinghamrisk score was preferred because it assigns higher risk with

diabetes and there are no age limits, and because it mayeasily be calculated in large cohorts. The European Heart-Score [24] is based on risk for patients between 40 and65 years and requires separate input for each individualusing charts or an online calculator, which was not prac-tical in our large cohort. To verify that the Framinghamrisk score was applicable in our cohort, we calculated thePC-based HeartScore for 50 randomly selected incidentRA cases and 50 controls. Parallel Framingham risk scoresand HeartScores were used to calculate an equation per-mitting estimation of the HeartScore in the entire cohortfrom their Framingham risk scores. An alternative logisticmodel was developed substituting the Framingham riskscore with these estimated HeartScores. In addition,we compared factors associated with incident cases ofself-reported RA with incident cases of self-reportedosteoarthritis.

ResultsThe baseline characteristics of the participants are pre-sented in Table 1. At baseline (1995 to 1997) 33,567 par-ticipants reported not having RA, and of this group 786(2.34%) reported RA at follow-up (2006 to 2008). Thiscorresponds to an average annual incidence of 0.21% inwomen and 0.16% in men, respectively. If restricted toparticipants with cases identified in the local hospitalregistries (n = 216), the annual incidence was 0.06% inwomen and 0.04% in men. As expected, incident casesof RA were older and more of them were current or pre-vious smokers and had hypertension at baseline. Inci-dent cases of RA had significantly elevated metabolicrisk factors including blood pressure, BMI, serum chol-esterol and triglyceride. The estimated median (inter-quartile range) Framingham risk scores at baseline were11 (7, 16) and 8 (4, 13), respectively, in incident casesand those who did not develop RA. Notably, a higherproportion of incident RA cases had a history of CV dis-ease at baseline.Female gender, age, smoking, BMI and previous CV

disease were more prominent in those developing RA(odds ratio 1.52 (1.11 to 2.07), Table 2, I - Main model).There was minimal change in the odds ratio for previousCV disease with inclusion of use of analgesics in this lo-gistic regression model, when systolic and diastolicblood pressure were included as continuous variables in-stead of the categorical variable for hypertension, withadditional adjustment for total cholesterol and HDL-cholesterol concentrations, or when a variable for phys-ical activity (low, moderate, high) was also included (datanot shown). In all the alternative multivariate models forincident RA where the number of patients was restrictedto better characterized subgroups, previous CV diseasewas significant and the odds ratios were higher than forthe main model including all self-reported cases (Table 2,

Table 1 Characteristics of individuals without rheumatoid arthritis (RA) or osteoarthritis (OA) at baseline

All participantswithout RA/OA

Later developedRA

Did not developRA

P-value Later developedOA

Did not developOA

P-value

n 33,567 786 32,781 3,586 29,981

Age1 (years) 46 (13) 51 (13) 46 (13) ** 52 (10) 45 (13) **

Sex (female) 18,207 (54.2%) 488 (62.1%) 17719 (54.1%) ** 2,406 (67.1%) 15,801(52.7%) **

Smoking ** **

Current 8,901 (26.5%) 256 (32.6%) 8,645 (26.4%) 1,015 (28.3%) 7,886 (26.3%)

Former smoker 7,908 (23.6%) 212 (27.0%) 7,696 (23.5%) 945 (26.4%) 6,963 (23.2%)

Never smoker 15,017 (44.7%) 281 (35.6%) 14,736 (45.0%) 1,425 (39.7%) 13,592 (45.3%)

Hypertension 12,186 (36.3%) 339 (43.1%) 11,847 (36.1%) ** 1,547 (43.1%) 10,639 (35.5%) **

Diabetes 454 (1.4%) 19 (2.4%) 435 (1.3%) * 61 (1.7%) 393 (1.3%)

Previous CV disease 1,060 (3.2%) 52 (6.6%) 1,008 (3.1%) ** 158 (4.4%) 902 (3.0%) **

Angina 673 (2.0%) 35 (4.5%) 638 (1.9%) ** 107 (3.0%) 566 (1.9%) **

MI 441 (1.3%) 16 (2.0%) 425(1.3%) 49 (1.4%) 392 (1.3%)

Stroke 223 (0.7%) 9 (1.1%) 214 (0.7%) 32 (0.9%) 191 (0.6%)

Family risk of CV disease 15,692 (46.7%) 414 (52.7%) 15,278 (46.6%) ** 2,017 (56.2%) 13,675 (45.6%) **

Framingham risk score2 9 (4 to 13) 11 (7 to 16) 8 (4 to 13) 11 (7 to 15) 8 (3 to 13)

Weight1 (kg) 76 (14) 77 (14) 76 (14) 77 (13) 76 (14)

BMI2 (kg/m2) 25.6 (23.5 to 28.1) 26.3 (24.3 to 29.0) 25.6 (23.4 to 28.1) ** 26.4 (24.2 to 29.0) 25.4 (23.3 to 27.8) **

Waist circumference1 (cm) 85 (11) 86 (11) 85 (11) * 86 (11) 85 (11) **

Hip circumference1 (cm) 102 (8) 103 (8) 102 (8) ** 103 (8) 101 (6) **

Waist/hip ratio1 0.84 (0.08) 0.84 (0.08) 0.84 (0.08) 0.83 (0.07) 0.84 (0.08) **

Systolic blood pressure2

(mmHg)131 (120 to 144) 133 (122 to 146) 131 (120 to 144) * 133 (122 to 147) 131 (120 to 143) **

Diastolic blood pressure1

(mmHg)79 (11) 80 (11) 79 (11) * 81 (11) 79 (11) **

Serum cholesterol1 (mmol/L) 5.8 (1.2) 6.0 (1.2) 5.8 (1.2) ** 6.1 (1.2) 5.8 (1.2) **

HDL cholesterol1 (mmol/L) 1.40 (0.39) 1.37 (0.39) 1.40 (0.39) 1.44 (0.40) 1.39 (0.38) **

Serum triglycerides2 (mmol/L) 1.40 (0.98 to 2.07) 1.51 (1.10 to 2.24) 1.40 (0.97 to 2.06) ** 1.46 (1.02 to 2.12) 1.37 (0.95 to 2.04) **1Values are mean (standard deviation) or numbers (%), or 2median (interquartile range). Non-respondents were excluded. BMI, body mass index; CVD, cardiovasculardisease; HDL, high density lipoprotein; MI, myocardial infarction; OA, osteoarthritis; RA, rheumatoid arthritis. *P-value <0.05, **P-value <0.001, comparing individuals whodeveloped RA and did not develop RA, or comparing individuals who developed OA and did not develop OA. All blood samples were non-fasting.


II – Alternative models A to F). A diagnosis of RA wasconfirmed in the hospital registries for 216 participants(27%) who had self-reported a new RA diagnosis inHUNT3. The characteristics of these participants werevery similar to those reported for the entire group of in-cident RA cases (data not shown). In the 216 cases witha hospital diagnosis of RA, the diagnosis was found oneto two years following enrolment in HUNT2 for 14cases (6.5%), after three years for 24 cases (11.1%), afterfour years for 15 cases (6.9%), after five years for 28cases (13.0%) and after six years for 22 cases (10.2%). Forthe remaining 113 cases (52.3%), the times were distributedfrom seven years onwards to the inclusion in HUNT3 withnumbers varying unsystematically from 15 to 24 cases(6.9 to 11.1%) per year. Thus, there was no obvious pat-tern regarding the number of years from HUNT2.

In a separate alternative multivariate logistic regressionmodel containing gender, previous CV disease and theFramingham risk score as covariates, the odds ratio fordeveloping RA in subjects with previous CV disease was1.65 (1.21 to 2.24) (P <0.01). This model also suggestedthe likelihood of developing RA was 6% greater witheach one unit increase in Framingham risk score (oddsratio: 1.06, 95% CI: 1.05 to 1.08), P <0.001). The modeldid not change with adjustment for BMI or daily use ofover-the-counter analgesics. In the model where theFramingham risk score was substituted with the estimatedHeartScore, previous cardiovascular disease remainedsignificant (odds ratio: 1.70 (1.25 to 2.32), P <0.01). In the100 participants where the HeartScore was calculated usingthe PC-based calculator, the HeartScore (logarithmicallytransformed) was highly correlated with the Framingham

Table 2 Effects of risk factors on incident RA and osteoarthritis, by multivariate logistic regression

Odds ratio 95% confidence interval P-value

I - Main model – Rheumatoid arthritis (n = 739 patients and 30,829 controls)1

Female gender 1.50 1.29 to 1.75 <0.01

Age2 1.03 1.02 to 1.04 <0.001

Current smoker 1.64 1.38 to 1.96 <0.001

Former smoker 1.32 1.10 to 1.59 <0.01

Hypertension 0.97 0.82 to 1.15 0.76

Diabetes 1.34 0.83 to 2.18 0.23

Body mass index2 1.04 1.02 to 1.06 <0.001

Previous CV disease 1.52 1.11 to 2.07 <0.01

II - Alternative models (sensitivity analyses) - Rheumatoid arthritis

A – After exclusion of patients with self-reported psoriasis (n = 573 patients and 30,829 controls)1

Previous CV disease3 1.70 1.23 to 2.36 0.001

B – Including patients with hospital diagnosis of RA (n = 201 patients and 30,829 controls)1


C – Including patients with hospital diagnosis of RA after 1999 (n = 178 patients and 30,829 controls)1

Previous CV disease3 2.26 1.26 to 3.99

D – Including patients with hospital diagnosis of RA after 2001 (n = 138 patients and 30,829 controls)1

Previous CV disease3 2.51 1.33 to 4.73 <0.01

E – Including patients carrying the Shared Epitope (n = 313 patients and 30,829 controls)1


F – Including patients carrying the Shared Epitope and excluding patients with self-reported psoriasis (n = 257 patients and 30,829 controls)1

Previous CV disease3 1.96 1.24 to 3.10 <0.01

III - Model for osteoarthritis (n = 3,364 patients and 24,631 controls)1

Female gender 2.36 2.15 to 2.56 <0.001

Age2 1.05 1.05 to 1.06 <0.001

Current smoker 1.42 1.30 to 1.56 <0.001

Former smoker 1.23 1.12 to 1.35 <0.001

Diabetes 0.90 0.66 to 1.22 0.49

Body mass index2 1.06 1.05 to 1.07 <0.001

Previous CV disease 1.04 0.86 to 1.27 0.661Cases with complete data.2Continuous variable, Odds ratio per one unit change.3Multivariate model including the same variables as the main model for RA. CV, cardiovascular; RA, rheumatoid arthritis.


risk score (R = 0.86, P <0.001). The correlation was verysimilar in patients (R = 0.89) and controls (R = 0.86).At follow-up (2006 to 2008), 3,586 participants had

developed osteoarthritis, corresponding to an averageannual incidence of 1.05% in women and 0.64% in men,respectively. These patients were older and the propor-tion of smokers, former smokers and of those withhypertension was higher than in the group that did notdevelop osteoarthritis. Systolic and diastolic blood pres-sure, BMI, hip and waist circumference, total cholesteroland triglycerides were higher for incident cases of osteo-arthritis, whereas HDL cholesterol was lower in incidentcases of osteoarthritis. Female gender, age and smoking

were associated with development of osteoarthritis, whereasprevious CV disease was not (Table 2 III).

DiscussionIn the present large population-based health study wefound that participants who developed RA betweenHUNT2 and HUNT3 had more CV disease and CV riskfactors prior to the onset of RA (at HUNT2) comparedto those without incident RA. A history of CV diseasewas a significant risk factor for incident RA, as well asthe previously reported risk factors: female gender, in-creasing age, increasing BMI and smoking [25]. Thefinding of a positive association with the Framingham


risk score may be related to smoking and age being in-corporated into the score. The supplementary analysisshowed that the Framingham score was equivalent to theestimated HeartScore in our population even though theFramingham score was not developed from European sub-jects. On the other hand, previous CV disease events didnot contribute to the risk of future osteoarthritis in ourstudy, indicating that the finding may be specific to RA.Smoking was associated with incident osteoarthritis.

CV disease and incident RAThe design of our study has several strengths. Since thedata on the CV risk factors and events were collected atHUNT2, recall bias at HUNT3 was greatly reduced com-pared to a design where patients report on these factorswhen receiving an RA diagnosis at a hospital-based clinic.Furthermore, the observation time between HUNT2 andHUNT3 was approximately 10 years. The importance of along observation time is underscored by the finding thatthe odds ratio for the association of incident RA with pre-vious CV disease increased when patients receiving a hos-pital diagnosis of RA during the first years followingHUNT2 were excluded. It also seems biologically plausiblethat longer exposure to increased inflammation due to CVdisease further increases the risk of incident RA. Further-more, the statistical power is greatly improved in apopulation-based cohort study due to the large number ofcontrols, and problems defining a relevant control groupas seen with case-control studies of RA are avoided.The major limitation of our study is that data were

self-reported. Previous studies suggest that self-reportedCV diagnosis and risk factors, such as hypertension andcigarette smoking, are reliable [26,27]. However, the highincidence rate of self-reported RA in our study com-pared, for example, to recent Swedish data based on in-patient and non-primary outpatient care (0.056% inwomen, 0.025% in men) [28], confirms that there prob-ably were many false-positive RA diagnoses. This wouldtend to decrease the power to detect true positive resultsin our study, but not bias towards false-positive findings.When including only cases identified in the hospitalregistries, the incidence was close to the Swedish data.Furthermore, the incidence of RA increases with obser-vation time [29]. Therefore, different data may not bedirectly comparable.Epidemiological studies also carry a risk of reverse

causation, which could ensue if persons with undiag-nosed RA in HUNT2 were erroneously included as inci-dent cases of RA, because RA in itself increases the riskof CV disease and events [10]. It seems unlikely that re-verse causation could explain our findings, given thatthe association with previous CV disease became stron-ger when the cases with shortest duration betweenHUNT2 and a hospital diagnosis of RA were removed

from the analysis. Furthermore, we did not observe atendency for a higher number of cases with a hospitaldiagnosis to occur during the earliest years followingHUNT2, which would have been expected with substan-tial reverse causation.While fatal CV events prior to the onset of RA also lim-

ited the participants in our study to survivors of thoseevents, this again would lead to under-estimation of theimpact of CV events on development of RA, or to restrictthe population at risk of RA to those ageing with CV riskthat was insufficient to lead to premature mortality. Auto-antibodies were not measured and seropositive and sero-negative cases of RA could not be distinguished, whichimpacts on the generalizability of our results compared toothers, such as population-based cohorts where it wasconcluded that ischemic heart disease was not increasedprior to RA onset [17]. There may also have been a selec-tion bias regarding participation in HUNT, where sickerpeople or those with a lower socioeconomic status are lesslikely to join. These people have a higher risk both for CVdisease and for RA, which may have biased our findings.The results from our various sensitivity analyses unani-

mously supported our main analysis. The overall conclu-sion that a history of CV disease is associated with the riskfor incident RA therefore seems reliable. Although ourstudy may not be suitable to give an exact estimate of thesize of this increased risk, the main analysis giving an ORof approximately 1.5 is probably conservative.

Risk factors for osteoarthritisOur results identifying smoking as a risk factor for fu-ture osteoarthritis contradict some previous studies,which report an inverse association between smokingand osteoarthritis [30]. However, the longitudinal pro-spective Clearwater study designed to identify risk factorsfor OA development, which included 2,505 participants,did not support this inverse association with smoking forany of four joint sites (knee, hand, foot or spine) [31].In another prospective study of 1,003 participants fromthe general population there was no association betweensmoking and radiologically-confirmed OA at differentsites [32]. The current study may have shown an asso-ciation with smoking because it is larger and includeda greater number of incident osteoarthritis diagnoses.Nevertheless, since our study relies on self-report, replica-tion in a cohort ascertained for radiographic osteoarthritisand symptoms is needed. However, in support of the ac-curacy of our case classification, we noted that higher BMIincreased the risk for osteoarthritis development, as previ-ously described [33].

Inflammation as a link between CV disease and incident RAFrom a biological perspective based on the current stud-ies, we hypothesize that inflammation may contribute to

Figure 1 Proposed model for illustrating the relationshipamong RA, CV disease, inflammation, atherosclerosis andpredisposing factors. CV, cardiovascular; RA, rheumatoid arthritis.


the parallel development of atherosclerosis and RA dur-ing the pre-clinical period in individuals exposed tocommon RA and CV risk factors, just as this interactionaccelerates complication of RA by CV events after onset.Inflammation is the most plausible link between previous

CV disease or BMI and increased risk for RA. It is wellestablished that the pathogenesis of CV disease, that is,atherogenesis, includes chronic inflammation of the wall ofmuscular arteries and that inflammation is increased duringacute coronary events, associated with unstable plaque andthrombosis [16]. Furthermore, obesity induces chronic in-flammation in adipose tissue, increasing pro-inflammatorycytokines including interleukin-6 and tumour-necrosisfactor [34]. In addition, a pro-inflammatory state is presentbefore the clinical onset of RA. C reactive protein (CRP)levels were found to be higher in individuals with pre-clinical RA compared to a control group [35]. Serum cyto-kines and chemokines were also elevated preceding theonset of RA [36]. In at least some patients, periodontal in-flammation may precede the onset of RA symptoms, alsoassociated with ACPA [37]. Respiratory inflammation hasalso been described in ACPA positive individuals withoutRA [38]. Autoantibodies may also increase the risk of im-mune complex-mediated vascular inflammation [39]. Thus,CV disease and CV risk factors increase the future risk ofRA development in the context of systemic inflammationand autoantibody development [12]. Patients with seroposi-tive RA are at greater risk of CV disease [6,40], and it ispossible that the RA genetic background is more permis-sive to atherosclerosis when associated with the “lifestyle”factors described here. In contrast to RA, in osteoarthritisinflammation is local and low-grade rather than systemic,and is not associated with autoantibodies.Figure 1 proposes a model for the relationships among

RA, CV disease, inflammation and various predisposingfactors, based on the results from the present study, as

well as previous studies demonstrating an increased riskof CV disease after diagnosis of RA. Given that our find-ings were robust when adjusting for known risk factorsfor RA, this hypothesis merits further investigation instudies designed to clarify pathogenetic mechanisms,which is not possible in an epidemiological study.

ConclusionsA history of previous CV disease was associated with in-creased risk of incident RA but not osteoarthritis. Aprobable explanation is that increased systemic inflam-mation may contribute to the parallel development ofatherosclerosis and RA during the pre-clinical period inindividuals exposed to common RA and CV risk factors,in a similar way that such interaction accelerates CV dis-ease in patients with established RA. The association ofprevious CV events with the development of RA at thepopulation level suggests presentation with CV events,especially in middle-aged female smokers or first-degreerelatives of RA patients, should raise clinical suspicionto capture cases of undiagnosed early RA. Because CVrisk factors are increased at the onset of RA, active car-diovascular risk management is important from the timeof diagnosis.

AbbreviationsACPA: Anti-citrullinated peptide antibodies; BMI: Body mass index;CV: Cardiovascular; HDL: High-density lipoprotein; HUNT: Nord-TrøndelagHealth Study; LDL: Low-density lipoprotein; MI: Myocardial infarction;OA: Osteoarthritis; RA: Rheumatoid arthritis.

Competing interestsThe authors declare that they have no competing interests.

Authors’ contributionsHP was responsible for conception and design, data collection and analysis,and manuscript writing. MAB contributed to conception and design, datacollection, and critical revision of the manuscript. SP contributed toconception and design, data analysis, and manuscript writing. RT took part inconception and design, and critical revision of the manuscript. VV wasresponsible for conception and design, data collection and analysis, andcritical revision of the manuscript. All authors read and approved the finalmanuscript.

AcknowledgementsThe HUNT study is a collaboration among the HUNT Research Centre(Faculty of Medicine, Norwegian University of Science and TechnologyNTNU), Nord-Trøndelag County Council, Central Norway Health Authority,and the Norwegian Institute of Public Health. We are grateful to Levanger,Namsos and Trondheim University hospitals for giving access to theirdiagnostic registries. This research was supported by NHMRC grant 569938,an Australian Postgraduate Award (HP) and an ARC Future Fellowship (RT).MAB is funded by an NHMRC Senior Principal Research Fellowship andQueensland Premier’s Fellowship.

Author details1University of Queensland Diamantina Institute, Translational ResearchInstitute, Princess Alexandra Hospital, Woolloongabba, QLD, Australia.2Queensland Clinical Trials and Biostatistics Centre, School of PopulationHealth, University of Queensland, Princess Alexandra Hospital,Wooloongabba, QLD, Australia. 3Department of Laboratory Medicine,Children’s and Women’s Health, Norwegian University of Science andTechnology, Trondheim, Norway. 4Department of Immunology andTransfusion Medicine, Trondheim University Hospital, Trondheim, Norway.


5Current address Clinical Trials & Biostatistics Unit, QIMR Berghofer MedicalResearch Institute, Herston, QLD, Australia.

Received: 16 October 2013 Accepted: 20 March 2014Published: 2 April 2014

References1. Maradit-Kremers H, Nicola PJ, Crowson CS, Ballman KV, Gabriel SE:

Cardiovascular death in rheumatoid arthritis: a population-based study.Arthritis Rheum 2005, 52:722–732.

2. Gonzalez-Juanatey C, Llorca J, Amigo-Diaz E, Dierssen T, Martin J, Gonzalez-Gay MA: High prevalence of subclinical atherosclerosis in psoriaticarthritis patients without clinically evident cardiovascular disease orclassic atherosclerosis risk factors. Arthritis Rheum 2007, 57:1074–1080.

3. Schonland SO, Lopez C, Widmann T, Zimmer J, Bryl E, Goronzy JJ, WeyandCM: Premature telomeric loss in rheumatoid arthritis is geneticallydetermined and involves both myeloid and lymphoid cell lineages.Proc Natl Acad Sci U S A 2003, 100:13471–13476.

4. Silman AJ, Newman J, MacGregor AJ: Cigarette smoking increases the riskof rheumatoid arthritis. Results from a nationwide study of disease-discordant twins. Arthritis Rheum 1996, 39:732–735.

5. Hazes JM, Dijkmans BA, Vandenbroucke JP, de Vries RR, Cats A: Lifestyleand the risk of rheumatoid arthritis: cigarette smoking and alcoholconsumption. Ann Rheum Dis 1990, 49:980–982.

6. Wallberg-Jonsson S, Johansson H, Ohman ML, Rantapaa-Dahlqvist S: Extentof inflammation predicts cardiovascular disease and overall mortality inseropositive rheumatoid arthritis. A retrospective cohort study fromdisease onset. J Rheumatol 1999, 26:2562–2571.

7. Sokka T, Hakkinen A, Kautiainen H, Maillefert JF, Toloza S, Mork Hansen T,Calvo-Alen J, Oding R, Liveborn M, Huisman M, Alten R, Pohl C, Cutolo M,Immonen K, Woolf A, Murphy E, Sheehy C, Quirke E, Celik S, Yazici Y,Tlustochowicz W, Kapolka D, Skakic V, Rojkovich B, Müller R, Stropuviene S,Andersone D, Drosos AA, Lazovskis J, Pincus T, QUEST-RA Group: Physicalinactivity in patients with rheumatoid arthritis: data from twenty-onecountries in a cross-sectional, international study. Arthritis Rheum 2008,59:42–50.

8. Toutouzas K, Sfikakis PP, Karanasos A, Aggeli C, Felekos I, Kitas G, Zampeli E,Protogerou A, Stefanadis C: Myocardial ischaemia without obstructivecoronary artery disease in rheumatoid arthritis: hypothesis-generatinginsights from a cross-sectional study. Rheumatology (Oxford) 2013, 52:76–80.

9. Koivuniemi R, Paimela L, Suomalainen R, Leirisalo-Repo M: Cardiovasculardiseases in patients with rheumatoid arthritis. Scand J Rheumatol 2013,42:131–135.

10. Liao KP, Solomon DH: Traditional cardiovascular risk factors, inflammationand cardiovascular risk in rheumatoid arthritis. Rheumatology (Oxford)2013, 52:45–52.

11. Del Rincon I, Williams K, Stern MP, Freeman GL, O'Leary DH, Escalante A:Association between carotid atherosclerosis and markers ofinflammation in rheumatoid arthritis patients and healthy subjects.Arthritis Rheum 2003, 48:1833–1840.

12. Hannawi S, Haluska B, Marwick TH, Thomas R: Atherosclerotic disease isincreased in recent onset rheumatoid arthritis: a critical role forinflammation. Arthritis Res Ther 2007, 9:R116.

13. Gonzalez-Juanatey C, Llorca J, Martin J, Gonzalez-Gay MA: Carotid intima-mediathickness predicts the development of cardiovascular events in patients withrheumatoid arthritis. Semin Arthritis Rheum 2009, 38:366–371.

14. Kremers HM, Crowson CS, Therneau TM, Roger VL, Gabriel SE: High ten-yearrisk of cardiovascular disease in newly diagnosed rheumatoid arthritispatients: a population-based cohort study. Arthritis Rheum 2008,58:2268–2274.

15. Kokkonen H, Soderstrom I, Rocklov J, Hallmans G, Lejon K, RantapaaDahlqvist S: Up-regulation of cytokines and chemokines predates theonset of rheumatoid arthritis. Arthritis Rheum 2010, 62:383–391.

16. McEntegart A, Capell H, Creran D: Cardiovascular risk factors, includingthrombotic variables, in a population with rheumatoid arthritis.Rheumatology 2001, 40:640–644.

17. Holmqvist ME, Wedren S, Jacobsson LT, Klareskog L, Nyberg F, Rantapaa-Dahlqvist S, Alfredsson L, Askling J: No increased occurrence of ischemicheart disease prior to the onset of rheumatoid arthritis: results from twoSwedish population-based rheumatoid arthritis cohorts. Arthritis Rheum2009, 60:2861–2869.

18. Holmen J, Midthjell K, Forsen L, Skjerve K, Gorseth M, Oseland A: TheNord-Trøndelag Health Study 1995–97 (HUNT2): Objectives, contents,methods and participation. Norsk Epidemiologi 2003, 13:19–32.

19. du Montcel ST, Michou L, Petit-Teixeira E, Osorio J, Lemaire I, Lasbleiz S,Pierlot C, Quillet P, Bardin T, Prum B, Cornelis F, Clerget-Darpoux F: Newclassification of HLA-DRB1 alleles supports the shared epitopehypothesis of rheumatoid arthritis susceptibility. Arthritis Rheum 2005,52:1063–1068.

20. Cortes A, Brown MA: Promise and pitfalls of the Immunochip. Arthritis ResTher 2011, 13:101.

21. Dilthey AT, Moutsianas L, Leslie S, McVean G: HLA*IMP—an integratedframework for imputing classical HLA alleles from SNP genotypes.Bioinformatics 2011, 27:968–972.

22. Walitt BT, Constantinescu F, Katz JD, Weinstein A, Wang H, Hernandez RK,Hsia J, Howard BV: Validation of self-report of rheumatoid arthritis andsystemic lupus erythematosus: the Women's Health Initiative.J Rheumatol 2008, 35:811–818.

23. D'Agostino RB Sr, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM,Kannel WB: General cardiovascular risk profile for use in primary care: theFramingham Heart Study. Circulation 2008, 117:743–753.

24. Conroy RM, Pyorala K, Fitzgerald AP, Sans S, Menotti A, De Backer G, DeBacquer D, Ducimetiere P, Jousilahti P, Keil U, Njølstad I, Oganov RG,Thomsen T, Tunstall-Pedoe H, Tverdal A, Wedel H, Whincup P, Wilhelmsen L,Graham IM; SCORE project group: Estimation of ten-year risk of fatalcardiovascular disease in Europe: the SCORE project. Eur Heart J 2003,24:987–1003.

25. Oliver JE, Silman AJ: Risk factors for the development of rheumatoidarthritis. Scand J Rheumatol 2006, 35:169–174.

26. Kehoe R, Wu SY, Leske MC, Chylack LT Jr: Comparing self-reported andphysician-reported medical history. Am J Epidemiol 1994, 139:813–818.

27. Martin LM, Leff M, Calonge N, Garrett C, Nelson DE: Validation ofself-reported chronic conditions and health services in a managed carepopulation. Am J Prev Med 2000, 18:215–218.

28. Eriksson JK, Neovius M, Ernestam S, Lindblad S, Simard JF, Askling J:Incidence of rheumatoid arthritis in Sweden: a nationwide population-based assessment of incidence, its determinants, and treatmentpenetration. Arthritis Care Res (Hoboken) 2013, 65:870–878.

29. Wiles N, Symmons DP, Harrison B, Barrett E, Barrett JH, Scott DG, Silman AJ:Estimating the incidence of rheumatoid arthritis: trying to hit a movingtarget? Arthritis Rheum 1999, 42:1339–1346.

30. Sandmark H, Hogstedt C, Lewold S, Vingard E: Osteoarthrosis of the kneein men and women in association with overweight, smoking, andhormone therapy. Ann Rheum Dis 1999, 58:151–155.

31. Wilder FV, Hall BJ, Barrett JP: Smoking and osteoarthritis: is there anassociation? The Clearwater Osteoarthritis Study. Osteoarthritis Cartilage2003, 11:29–35.

32. Hart DJ, Spector TD: Cigarette smoking and risk of osteoarthritis inwomen in the general population: the Chingford study. Ann Rheum Dis1993, 52:93–96.

33. Kulie T, Slattengren A, Redmer J, Counts H, Eglash A, Schrager S: Obesityand women's health: an evidence-based review. J Am Board Fam Med2011, 24:75–85.

34. Tuttle HA, Davis-Gorman G, Goldman S, Copeland JG, McDonagh PF:Proinflammatory cytokines are increased in type 2 diabetic women withcardiovascular disease. J Diabetes Complications 2004, 18:343–351.

35. Masi A, Aldag J, Sipes J: Do elevated levels of serum C-reactive proteinpredict rheumatoid arthritis in men: correlations with pre-RA status andbaseline positive rheumatoid factors. J Rheumatol 2001, 28:2359–2361.

36. Hughes-Austin JM, Deane KD, Derber LA, Kolfenbach JR, Zerbe GO,Sokolove J, Lahey LJ, Weisman MH, Buckner JH, Mikuls TR, O'Dell JR, KeatingRM, Gregersen PK, Robinson WH, Holers VM, Norris JM: Multiple cytokinesand chemokines are associated with rheumatoid arthritis-relatedautoimmunity in first-degree relatives without rheumatoid arthritis:Studies of the Aetiology of Rheumatoid Arthritis (SERA). Ann Rheum Dis2013, 72:901–907.

37. Chen H-H, Huang N, Chen Y-M, Chen T-J, Chou P, Lee Y-L, Chou Y-J, Lan J-L,Lai K-L, Lin C-H, Chen DY: Association between a history of periodontitis andthe risk of rheumatoid arthritis: a nationwide, population-based,case–control study. Ann Rheum Dis 2013, 72:1206–1211.

38. Fischer A, Solomon JJ, du Bois RM, Deane KD, Olson AL, Fernandez-PerezER, Huie TJ, Stevens AD, Gill MB, Rabinovitch AM, Lynch DA, Burns DA,


Pineiro IS, Groshong SD, Duarte Achcar RD, Brown KK, Martin RJ, Swigris JJ:Lung disease with anti-CCP antibodies but not rheumatoid arthritis orconnective tissue disease. Respir Med 2012, 106:1040–1047.

39. Kerekes G, Szekanecz Z, Dér H, Sándor Z, Lakos G, Muszbek L, Csipö I, SipkaS, Seres I, Paragh G, Kappelmayer J, Szomják E, Veres K, Szegedi G,Shoenfeld Y, Soltész P: Endothelial dysfunction and atherosclerosis inrheumatoid arthritis: a multiparametric analysis using imagingtechniques and laboratory markers of inflammation and autoimmunity.J Rheumatol 2008, 35:398–406.

40. Goodson N, Marks J, Lunt M, Symmons D: Cardiovascular admissions andmortality in an inception cohort of patients with rheumatoid arthritiswith onset in the 1980s and 1990s. Ann Rheum Dis 2005, 64:1595–1601.

doi:10.1186/ar4527Cite this article as: Pahau et al.: Cardiovascular disease is increased priorto onset of rheumatoid arthritis but not osteoarthritis: the population-based Nord-Trøndelag health study (HUNT). Arthritis Research & Therapy2014 16:R85.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

cardiovascular disease in rheumatoid arthritis367696/s...cardiovascular disease in rheumatoid...

Documents