A PILOT STUDY OF THE PREVALENCE OF COELIAC DISEASE IN VIETNAMESE PATIENTS WITH GRAVES’ DISEASE

RESULTS 35 4.1 Recruitment of Participants………..35 4.2 Biochemical Analysis………..35 TABLE I: Baseline Characteristics of the Participants……….35 TABLE II: Prevalence of coeliac disease by

HONOUR THESIS

A PILOT STUDY OF THE PREVALENCE OF COELIAC DISEASE IN

VIETNAMESE PATIENTS WITH GRAVES’ DISEASE

Khoa Dang Truong, BBMed Sci

207624

Supervisors

Professor Peter R Ebeling, MD, MBBS, FRACP

Dr Alex Boussioutas, MBBS, PhD, FRACP

Dr Bob Anderson, PhD, FRACP

Thesis submitted to University of Melbourne for the Degree of

Bachelor of Science with Honours

Department of Medicine (RMH/WH) Monday 27th October 2008 Word Count: 8,969

DEDICATION

~ This Honours Research Project and Thesis is dedicated to my Parents ~

In remembrance of my grandfathers Truong Thuong and Le Duc Thanh, both of whom had passed

away with cancers

~ Their memories will remain with me for as long as I live ~

DECLARATION BY PRINCIPLE SUPERVISOR

I confirm that the declaration above of (Student’s Name) Thesis is a true and fair representation of the student’s work

Khoa Dang Truong

Signature: _ Date: _23/10/2008

*Your Acknowledgement page must declare, as appropriate:

The extent to which the student has used the work of others

The contribution of the student to work carried out in collaboration with others

A description of work submitted for any other qualification

A description of work carried out prior to enrolment in Honours

*Your Acknowledgement page must declare, as appropriate:

The extent to which the student has used the work of others

The contribution of the student to work carried out in collaboration with others

A description of work submitted for any other qualification

A description of work carried out prior to enrolment in Honours

*Your Acknowledgement page must declare, as appropriate:

The extent to which the student has used the work of others

The contribution of the student to work carried out in collaboration with others

A description of work submitted for any other qualification

A description of work carried out prior to enrolment in Honours

ACKNOWLEDGEMENTS

“Trust in the LORD with all your heart and lean not on your own understanding;

In all your ways acknowledge him, and he will make your paths straight.” ~ Proverbs 3:5-6

In reverence of the Lord God Almighty, I am eternally grateful for His providence, blessings and faithfulness in all things Let all glory be to God, forever and ever Amen

To my supervisor, Professor Peter Ebeling, thank you for your gentle words, your understanding and compassion Your unfathomed wisdom had always remained a mystery to me, even to this day Thank you for giving me a chance to follow my dreams and to experience research that would benefit others To my co-supervisor, Dr Alex Boussioutas, thank you for your generous guidance, patience and lenience Thank you for reminding me the reason why I worked so hard to be where I am today You will always have my respect, both as a scientist and a man of his word To my co-supervisor, Dr Bob Anderson, thank you for your genuine advice, reassurance and unrivalled expertise You were always there for me when there was no one else to turn to I will always remember the motivational chats and the meetings over lunch with you To all three of you, thank you for making this the best year of my study-life

To Dr Claudia Gagnon, thank you for the time you have spent, taking me through the ethics approval process and showing me the statistical analysis I can never repay you enough for your kind intentions, and for helping me along every step of the way in my Honours Project this year To Dr Anjali Haikerwal, thank you for your references regarding the ranges of Vitamin D deficiency in the project To Dr Leo Rando, thank you for your referrals of patients from Endocrinology clinic

To Sandy Dupuis, thank you for building and managing the database of participants for this project Your wonderful ideas and constructive criticisms were necessary and influential in the arrangements of this project To Maria Bisignano, thank you for directing the blood tests at Pathology

of Royal Melbourne Hospital You were so dedicated and thoughtful, which made everything so easy

to overcome To Margaret Watson, thank you for organising the blood collection at Western Hospital Even though you were short on staff, you still maintained an easygoing attitude and tolerance towards the rapidly increasing number of participants who came for blood test To Cathy Phillips, thank you for standing up for me in times of need Your presence alone, at the Pathology office, had encouraged

me to endure through all the tough times To all the nurses who helped with the blood collection, thank you for your impressive skills and the ability to work under pressure

To Marian Croft, thank you for your suggestions regarding the blood tests and patient history

To Sally Edmonds, thank you for your directions with the online clinical information system To Jia Wei Teo, thank you for the transport and storage of the frozen blood samples in this project To Vivien Li, thank you for your fluent writing skills that were used to edit my ethics approval forms To Lynette Kalms, thank you for general administration and receiving patient phone calls in my absence

To Fiona Wylie, thank you for your encouragement and warm-hearted smiles It wouldn’t be the same without you around the department To Erin Wijaya Ng, thank you for the initial accommodations of the project I wish you all the best in your future career

To my cousin, Nguyen Tuong Dinh, thank you for your reading of the daily radio advertisements on Stereo FM 97.4 Vietnamese Radio Your specialty in radio advertisement had predisposed the potential participant to participation To Mr Thai Hoa and Mrs Phuong Hoang from SBS Vietnamese Radio 122.4 AM, thank you for the valuable time of your interview and daily advertisement broadcasts It was terrific to have you communicated our project to the broader Vietnamese Community To Dr Hai Phan, from the local clinics, thank you for your referral of participants and offer to promote our project to the Vietnamese Medical society To the Vietnamese Churches and Organisations, thank you for your overwhelming support for this project

To my Honours friends and fellows DoMSA Committee who have travelled with me on this journey, thank you for your companionship and your support in my role as Honours Student Representative this year in 2008 I will never forget the precious time spent together on our Ski Trip, rock climbing, soccer games, and trivia nights Thank you all for being part of my life

To my grandmother, Nguyet Thi Le, thank you for praying to God with tears for me To aunt Dung Thi My Truong, thank you for your enthusiasm in contacting friends and neighbours To my Father, Luu Vinh Truong; my mother, Hoa Kim Le; and my beloved brothers, Danh Cong Truong, Long Kim Truong and Thomas An Thien Truong Thank you You are the reason why I am here

To all Vietnamese participants whose participation had made this Project possible… Thank you!

TABLE OF CONTENTS

TITLE PAGE 1

DEDICATION 2

DECLARATION 3

PAGE 1 OF 2……….……….3

PAGE 2 OF 2……….……….4

ACKNOWLEDGEMENTS 5

TABLE OF CONTENTS 7

ABSTRACT 10

ABBREVIATIONS 12

1 INTRODUCTION 14

1.1 GRAVES’ DISEASE……… 14

1.1.1 Graves’ disease as a Cause of Hyperthyroidism……… 14

1.1.2 The Aetiology of Graves’ disease – Role of Autoimmunity……… 14

1.1.3 Genetic Background……….15

1.1.4 Prevalence of Graves’ disease……….16

1.1.5 Diagnosis of Graves’ disease……… 16

1.1.6 Current Therapy and Treatment……… 17

1.2 COELIAC DISEASE……… 17

1.2.1 Coeliac disease as an Autoimmune Disorder……… 17

1.2.2 The Aetiology of coeliac disease – Disease Mechanism and Cause……… 17

1.2.3 Genetic Background……….19

1.2.4 Prevalence of coeliac disease……… 20

1.2.5 Symptoms and Complications……….20

1.2.6 Diagnosis and Screening……… 21

1.2.7 Current Therapy and Treatment……… 22

1.2.8 Consequences of untreated coeliac disease……… 23

1.3 RATIONALE……… 24

1.3.1 The Link between Graves’ and coeliac disease……… 24

1.3.2 High rate of Vietnamese Graves’ patients at Western Hospital……… 24

1.3.3 Early Prevention reduces Morbidity, Mortality and other Risks……….25

1.3.4 Micronutrient Deficiencies Might Predict coeliac disease……… 25

1.3.5 To Establish the Prevalence of coeliac disease in Vietnamese………25

2 AIMS 26

2.1 Hypothesis………26

2.2 Primary Aim……….26

2.3 Secondary Aim……….26

3 METHODS 27

Ethics Approval……… 27

3.1 PROJECT DESIGN……….27

3.1.1 Study Population……… 27

3.1.2 Recruitment Strategies……….27

3.1.3 Inclusion Criteria……….31

3.1.4 Exclusion Criteria………31

3.1.5 Database Establishment……… 32

3.2 PROCEDURES………32

3.2.1 Initial Screening……… 32

3.2.2 Study Visit……… 32

3.2.3 Biochemical Blood Tests……….32

3.2.4 Genes Test and Small Bowel Biopsy……… 33

3.3 STATISTICAL ANALYSIS……… 33

3.3.1 Sample size and Selection………33

3.3.2 SPSS Software Analysis……… 33

3.3.3 Limitations……… 34

3.3.4 Potential Source of Bias……… 34

4 RESULTS 35

4.1 Recruitment of Participants……… 35

4.2 Biochemical Analysis……… 35

TABLE I: Baseline Characteristics of the Participants……….35

TABLE II: Prevalence of coeliac disease by Group……….36

TABLE III: Prevalence of Vitamin D Deficiency and Insufficiency by Group……… 36

FIGURE 1: Prevalence of Severe, Moderate and Mild Vitamin D Deficiency and Insufficiency by Group……….37

TABLE IV: Prevalence of other Nutrient Deficiencies by Group………37

FIGURE 2: Prevalence of other Nutrient Deficiencies by Group………38

TABLE V: Prevalence of Anaemia, Microcytosis and Macrocytosis by Group……… 38

FIGURE 3: Prevalence of Anaemia, Microcytosis and Macrocytosis by Group……….39

5 DISCUSSION 40

5.1 Baseline Characteristics……… 40

5.2 Prevalence of coeliac disease in Vietnamese Graves’ disease Patients……… 40

5.3 Vitamin D Deficiency……… 41

5.4 Anaemia in Graves’ disease……….42

5.5 Limitations Encountered……… 42

5.6 Source of Bias……… 42

6 CONCLUSION 44

7 REFERENCES 45

APPENDIX i

APPENDIX 1: PARTICIPANT INFORMATION AND CONSENT FORM English Version………i

Consent Form……….vi

Revocation of Consent Form………vii

APPENDIX 2: PARTICIPANT INFORMATION AND CONSENT FORM Vietnamese Version……….viii

Consent Form……… xiii

Revocation of Consent Form……… xiv

APPENDIX 3: PARTICIPANT QUESTIONNAIRES – English Version Patient Questionnaires……… xv

SF36 Health Survey………xvii APPENDIX 4: PARTICIPANT QUESTIONNAIRES – Vietnamese Version

Patient Questionnaires………xxii

SF36 Health Survey……… xxiv

APPENDIX 5: INVITATION LETTER TO PARTICIPANTS – English Version……… xxix

APPENDIX 6: INVITATION LETTER TO PARTICIPANTS – Vietnamese Version……….xxx

APPENDIX 7: ADVERTISEMENT POSTER – English Version………xxxi

APPENDIX 8: ADVERTISEMENT POSTER – Vietnamese Version……….xxxii

APPENDIX 9: DE-IDENTIFIED CODE KEY……… xxxiii

APPENDIX 10: PATHOLOGY REFERENCES……….xxxiv

APPENDIX 11: PATIENT QUESTIONNAIRE ANSWERS PART 1 OF 3………xxxv

PART 2 OF 3……… xxxix PART 3 OF 3……… xliii APPENDIX 12: SF36 HEALTH SURVEY ANSWERS

PART 1 OF 2……….xlvii PART 2 OF 2……… li APPENDIX 13: PARTICIPANT BLOOD TEST RESULTS

PART 1 OF 2……….lv PART 2 OF 2………lix

ABSTRACT

Current data suggest that the prevalence of coeliac disease is high in Caucasian with Graves’ disease However, there are no data assessing the prevalence of coeliac disease in Vietnamese with Graves’ disease; nor data regarding micronutrient deficiency in Graves’ disease patients, and whether this is higher compared with the normal Vietnamese population

In this study, we aimed to establish the prevalence of coeliac disease in Vietnamese patients with Graves’ disease The prevalence was compared with Vietnamese participants without autoimmune disease, and also with the prevalence in Caucasians as previously reported in the literature Furthermore, the frequency of micronutrient deficiency was determined in Vietnamese patients with Graves’ disease and in control participants

The population of participants included 60 Vietnamese Graves’ disease patients and 120 Vietnamese control participants, who were recruited using various recruitment strategies The most successful strategy involved Radio Advertisement and a national-broadcasted interview on two well-known Vietnamese radio stations, Stereo FM 97.4 Vietnamese Radio and SBS Vietnamese Radio 122.4 AM Consequently, the outcome of recruitment has shown an effective study design needs to overcome the underlying cultural misconceptions and implications, to investigate a range of useful methods that successfully accessed the whole Vietnamese community

Altogether, 180 Vietnamese participants were invited, who gave consent to answer two lifestyle questionnaires, and underwent a collection of blood sample for pathology testing The results were organised into a database and statistically analysed, using different combinations of unpaired Student t-test, Mann Whitney test, Pearson’s Chi-square and Fisher’s exact test Coeliac disease was diagnosed using IgA anti-tTGA test in 1 Vietnamese Graves’ disease patient and 2 Vietnamese control participants (excluding 1 with borderline result) Thus, the calculated prevalence of coeliac disease in Vietnamese Graves’ patients is 1:60, which is similar to Vietnamese control participants; but likely to be lower, compared to 1:22 prevalence of coeliac disease found in Caucasian Graves’ disease patients

On the other hand, we observed a high prevalence of Vitamin D deficiency in the Vietnamese participants of our pilot study The rate of Vitamin D deficiency and insufficiency was found to be

88.3%, while Vitamin D deficiency alone contributed to 55.8% of the Vietnamese population in the study We speculated the involvement of environmental factors, including seasonal exposure, low dietary intake of Vitamin D, lifestyle factors and geographical latitude of southern Australia, which might be important given the emerging public health issue of Vitamin D deficiency in Australia Other micronutrient deficiencies observed were not significantly different between groups

Future larger studies are required to confirm the prevalence of coeliac disease in Vietnamese Graves’ disease patients, and investigate the high prevalence of Vitamin D deficiency found in this pilot study The results obtained will lay the foundation for power and sample size calculation for larger population-based studies, which should also confirm the HLA-DQ type in the Vietnamese population

CTLA4 Cytotoxic T-Lymphocyte Antigen 4

FBE Full Blood Examination

HCO 3 Bicarbonate Ion

HLA Human Leucocyte Antigen

HREC Human Research Ethics Committee

IgA Immunoglobulin Type A

MCH Mean Corpuscular Haemoglobin

MCHC Mean Corpuscular Haemoglobin Concentration

MCV Mean Corpuscular Volume

MHC Major Histocompatibility Complex

MPV Mean Platelet Volume

Na Sodium

PCV Packed Cell Count

PICF Participant Information and Consent Form

PLT Platelets

RCC Red Cell Count

RDW Red Cell Distribution Width

RMH Royal Melbourne Hospital

SBB Small Bowel Biopsy

SD Standard Deviation

sFol Serum Folate

T3 Triiodothyronine

T4 Thyroxine (also L-Tetraiodothyronine)

TPA Thyroid Peroxidase Antibodies

TRAbs Thyrotropin Receptor Antibodies

TSH Thyroid-Stimulating Hormone (also Thyrotropin)

TSHR Thyrotropin Receptor

TSH-RAbs Thyrotropin Receptor Antibodies

TSI Thyroid-Stimulating Immunoglobulin

1 INTRODUCTION

1.1 GRAVES’ DISEASE

1.1.1 Graves’ disease as a Cause of Hyperthyroidism

Graves’ disease (GD) is the most common cause of hyperthyroidism It is characterised by hyperthyroidism (overactive thyroid gland) [1], goitre, exophthalmos (ophthalmopathy) and rarely pretibial myxoedema (dermopathy) It is a condition where the thyroid gland, triggered by endogenous auto-antibodies acting like Thyroid-Stimulating Hormone (TSH), produces excess thyroid hormones that ultimately affect the overall body metabolism

In normal individuals, the thyroid gland produces and secretes thyroxine (T4), and in smaller quantities, triiodothyronine (T3) Both hormones control a range of bodily functions such as heart rate, appetite and body temperature The serum levels of these hormones act as negative feedback for the pituitary gland in the brain In hyperthyroidism, high serum levels of free T4 and T3 are sensed by the pituitary gland and consequently, the release of TSH is decreased

1.1.2 The Aetiology of Graves’ disease – Role of Autoimmunity

GD has been suggested to arise from a complex interaction between susceptibility genes and environmental modulating factors (e.g dietary iodine) [2] Epidemiologic evidences revealed that an increased risk of GD is found in those whose family members were diagnosed with the disease [3], as well as in individuals having other immune diseases The incidence of GD was also reported to be in excess of five to ten folds in women compared to men [4, 5]

In patients with Graves’ disease, the body’s immune system mistakenly attacks its own healthy cells and body tissues This autoimmunity involves the initial production (by unknown trigger)

of autoantibodies called Thyroid-Stimulating Immunoglobulin (TSI) by the immune system Other names for TSI include TSH-receptor antibodies (TRAbs or TSH-RAbs), thyroid-binding inhibiting immunoglobulin and long-acting thyroid stimulator Once produced, TSI binds to thyrotropin receptors (TSHR) on thyroid follicular cells, to mimic the action of TSH and stimulates T4 and T3 production in the thyroid gland When excess T4 is produced due to over-activation and is chronically secreted into the bloodstream, it affects many different body systems and gives rise to an extensive

range of serious symptoms [6] Unfortunately, TSI is not subjected to negative feedback of the pituitary gland Therefore, it continues to be produced and causes autonomous stimulation of TSHR This leads to high levels of thyroid hormones being released, as well as thyroid growth (causing a diffuse goitre) The individual then becomes hyperthyroid

GD could be associated with other auto-immune diseases including pernicious anaemia [7], vitiligo [8], myasthenia gravis [9], Sjögren’s syndrome [10], rheumatoid arthritis [11], lupus [12] and idiopathic thrombocytopenic purpure [13]

1.1.3 Genetic Background

Genetic factors are responsible for 79% of the susceptibility to develop GD [14], while environmental factors contribute to the remaining 21% [15] The first genetic factor found (in Caucasians) to be in conjunction with an increased risk of GD was HLA-B8 [16], a Human Leucocyte Antigen of Class I Major Histocompatibility Complex (MHC) It has relative risks ranging from 1.5 to 3.5 [17] Later, a stronger association was discovered between GD and HLA-DR3 [18], which is now known to be in linkage disequilibrium with HLA-B8 [19] The frequency of HLA-DR3 in the general population is approximately 15 to 30% compared with 40 to 55% in GD patients, resulting in a relative risk ranging from 3 to 4 [20, 21] In addition, the association between HLA-DQA1*0501 and

GD was also shown in Caucasians, with relative risk of 3.8 [22-24] However, the primary susceptibility allele remains to be HLA-DR3 (HLA-DRB1*03), as was suggested by recent studies [25] Other genetic susceptibility factors were also found in different ethnic groups including Chinese (HLA-Bw46) [26-30], Japanese (HLA-B35) [31, 32] and Korean (HLA-DR5) [33] There are no known genetic susceptibility factors in Vietnamese

Research on non-HLA is currently unclear and identification of other gene loci may require more extensive analysis on the entire human genome Several reports have defined a non-HLA gene that encodes for the negative co-stimulatory molecule cytotoxic T-lymphocyte antigen 4 (CTLA4), but the described effect of this gene in GD is quite small [34-38]

1.1.4 Prevalence of Graves’ disease

There is no study that has reported the prevalence for GD in the general population Nevertheless, hyperthyroidism was found to affect up to 2% of the population [4, 5, 39-42] It is logical to reason that, since GD affects approximately 60 to 80% of all patients with hyperthyroidism [43], the prevalence of GD would be less than 1.6% of the whole population In other words, the prevalence could be as high as 1:63 The distribution of GD around the world appears to be equal, moderately affecting all countries and races [44] In any case, the prevalence was reported to be similar among Caucasians and Asians, but found to be lower among Sub Saharan African descent [45]

In the United States, it was claimed that GD is the most prevalent autoimmune disorder [46]

1.1.5 Diagnosis of Graves’ disease

The diagnosis of GD is often performed by an endocrinologist or doctor, based on the presenting clinical symptoms, laboratory evaluation and imaging studies

Once GD is suspected from clinical diagnosis, laboratory data are then required to verify the diagnosis, estimate the severity of condition and assist in therapy planning The typical tests involved are serum TSH, free T4 and T3 levels If the results show high free T4 and T3 levels and low TSH, it would indicate that excess thyroid hormones are being produced and secreted, while TSH is under suppression by the negative feedback mechanism Considering this outcome in conjunction with clinical examination, GD would be diagnosed However, if the results turn out to be inconclusive, two additional blood tests might be contemplated, which includes thyroid-peroxidase antibodies (TPA) test and TSI antibodies assay [47] It was found that up to 90% of GD patients have TPA, previously known as ‘microsomal antigen’ [44] To a lesser extent are conventional radioreceptor antibody assays (RAAs), which would be able to detect TRAbs in 80 to 90% of the cases Therefore, a TPA test is effective in verifying GD diagnosis However, following the advent of second generation TRAbs assays with higher sensitivity of up to 99% [48], the use of these assays had superseded the TPA test

1.1.6 Current Therapy and Treatment

There is no cure for GD However, the disease can be treated with radioiodine therapy, antithyroid drugs or thyroid surgery, often followed by a life-long medication regime (thyroid hormone replacement therapy) Preferences for each mode of treatment vary according to circumstances and regions, with different degrees of effectiveness and potential side effects All three methods had together provided satisfactory outcome in over 90% of GD patients [49]

1.2 COELIAC DISEASE

1.2.1 Coeliac disease as an Autoimmune Disorder

Coeliac disease (CD) is an immune disorder of the small intestine characterised by mucosal inflammation, villous atrophy and crypt hyperplasia [50] It is triggered by exogenous gluten (protein

in wheat, barley and rye grains) and ultimately results in small bowel inflammation and malabsorption [51]

In addition, CD is classified as a delayed-type hypersensitivity reaction Other delayed-type hypersensitivity diseases include chronic transplant rejection [52] and atopic dermatitis (eczema) [53] Furthermore, CD associated with other diseases [54] including type I diabetes [55], Down’s syndrome [56], Addison’s disease [57], Sjögren’s syndrome [58], osteopenia [59], anaemia [60] and irritable bowel syndrome [61] On the other hand, dermatitis herpetiformis is currently considered a cutaneous manifestation of CD rather than an associated disease [62]

1.2.2 The Aetiology of coeliac disease – Disease Mechanism and Cause

CD is a multifactorial systemic disorder Hence, it was suggested that CD development involves the correlated action between HLA and non-HLA genes, along with gluten and other environmental regulating factors [63] Further evidence pointed out that the cause for an increased risk of CD is found in individuals with a family history of the disease [64], as well as those of European descent or having other immune diseases For unknown reasons, CD occurs in two to three times as many women as it is in men [65] CD can be manifest and diagnosed at any age However, it

was noted that the onset of disease based on immunological markers is nearly always within the first five years of life [66]

It is important to note that immunogenic peptides present exclusively in dietary gluten are the responsible stimulus of CD pathogenesis These peptide fragments are proline- and glutamine-rich protein from the gliadin and glutenin fraction of gluten [67] Strictly speaking, gluten only covers those disease-activating peptides in wheat [68] In rye and barley, the secalins and hordeins closely related to wheat gluten, are toxic in CD Other grains related to bread-making wheat, rye and barley that contain pathogenic peptides include triticale, spelt and their derivatives In contrast, the analogous proline- and glutamine-rich seed-storage proteins and peptides found in rice, maize (zein in corn), millet and sorghum do not trigger CD Up until now, it remains a highly controversial issue whether oats (avein) would be safe for consumption in CD patients [69]

The ingestion of gluten (and related prolamines found in wheat, rye, barley and possibly oats)

in some of those genetically predisposed individuals eventually leads to inflammation and damage of the small intestinal mucosa This occurs in sequence, starting with the incomplete digestion of gluten

by gastric, duodenal, pancreatic enzyme and brush-border membrane proteases; breaking it down into relatively protease-resistant peptide, for example a 33-amino-acid (33-mer) peptide molecule [70] By itself and in its native state, the 33-mer peptide molecule directly acts as a potent activator of specific T-cell lines in CD [71], which causes T-cell proliferation and subsequent intestinal toxicity Alternatively, being a gliadin peptide, 33-mer constitutes an excellent substrate for the ubiquitous tissue transglutaminase (tTG) enzyme [72] Thus, selective enzymatic deamidation of this peptide molecule by tTG increases its affinity of binding to HLA-DQ2 cell-surface antigens [73]; which can lead to priming and maintenance of an inflammatory T-cell response and enteritis Both of these routes of activation require that the gliadin peptide either be absorbed via intercellular tight junctions

or transcellular routes into the lamina propria, before being presented by antigen-presenting cells to CD4+ T-cells [74] Once they become activated, T-cells signal other immune and inflammatory cells

to produce cytokines and tissue destruction, also providing help to B-cells to generate antibodies specific for gluten and a variety of autoantigens Intraepithelial lymphocytes are also activated and are likely to be critical to epithelial damage and the malabsorption characteristic of CD

1.2.3 Genetic Background

CD is predominantly a disease of Caucasians [75] The first genetic factor found to be in conjunction with an increased risk of CD was HLA-B8 and HLA-A1 [76, 77] Later on, a stronger association was discovered between CD and HLA-DR3 (i.e HLA-DRw3 is the alloantigen of HLA-Dw3) [78], as well as HLA-DQ2 [79-81] Both HLA-DR3 and DQ2 are in strong linkage disequilibrium with each other and also HLA-A1 and B8 as a defined haplotype, relatively specific for CD [82, 83] Individually, HLA-DQ2 was found in more than 90% of CD patients, compared with

20 to 30% of healthy controls [84] In the absence of HLA-DQ2, the association between HLA-DQ8 and CD was also observed [85] Hence, the primary susceptibility was conclusively determined to be the consequence of both peptide-presenting molecules HLA-DQ2 (DQA1*05 and B1*02) and HLA-DQ8 (DQA1*0301 and B1*0302) [86] It was claimed that, together HLA-DQ2 and HLA-DQ8 are found in more than 99% of CD patients, compared with 40% of the general population [87] One report had suggested HLA-DQ8 to be of appreciable frequency in the Japanese population, while HLA-DQ2 is quite rare [88] However, this report went on to say that CD is unknown in Japan, mainly because the diet there is rice instead of wheat Another report claimed that the incidence of CD

in Asian children is as high as in Caucasian children [89] This report also went on to say that anecdotal evidence claims that CD ‘can be seen’ in Chinese children It is therefore, ‘not clear whether Asian populations actually have a genetic risk for CD’ After all, CD has been said to be rarely diagnosed in Chinese, Japanese or Africans (although some cases have been found in Asians from India and Pakistan) [74, 90, 91]

Less is known about the non-HLA susceptibility gene of CD, which is also the case with other polygenic inflammatory diseases It was hypothesized that non-HLA genes have limited effect on the pathogenesis of CD [73] Nevertheless, a region on the long arm of chromosome 5 (5q31-33) was found to be consistent with CD [92] In addition are some evidences for susceptibility factors on chromosome 11q [93] and 19p13 [94] Several reports have also defined non-HLA polymorphisms in CTLA4 genes that are most consistent with GD [95, 96]

1.2.4 Prevalence of coeliac disease

According to many reports, the prevalence of CD could be as high as 1:100 [97-101] CD prevalence was found to range from 0.5 to 1%, as the result of serological screening in “healthy” volunteers around the world [102], which is mostly from Europe, South America, Australia and the United States [103] In Australia, the prevalence was recently reported to be 1:251 [104, 105] The highest reported prevalence is observed in Western Europe and also in places where Europeans have emigrated [74] Since CD was considered to be predominant in Caucasian populations and observed

to be uncommon amongst oriental Asians, the prevalence was not established in many Asian countries

1.2.5 Symptoms and Complications

The classical definition of CD includes mucosal inflammation, with verifications of villous atrophy and crypt hyperplasia confirmed by small bowel biopsy (SBB), and recovery of villous architecture in response to gluten-free diet (GFD) Other bowel symptoms include those of malabsorption, weight loss and signs of nutrient or vitamin deficiency [106] New definitions of CD may involve different combinations of clinical examination, serology test, genetic determination and SBB indications

The lesion in CD, which might be found by SBB, is localized in the proximal part of the small intestine [63] For individual cases of diagnosed CD, the severity would be determined depending on where it falls in the spectrum of pathology, according to Marsh’s three stages of classification: the infiltrative (mucosal inflammation), the hyperplastic (crypt hyperplasia), and the destructive lesions (villous atrophy) [107] Later on, these morphologic findings were codified to produce the modified Marsh classification [108]

Major advances have defined CD to be more of a multisystem immunological disorder, rather than simply being a gastrointestinal disease characterised by malabsorption and diarrhoea [109] Severe symptoms that commonly appear in early childhood include chronic diarrhoea, abdominal distension, and failure to thrive Meanwhile, many patients may not develop any symptom until later

in their adult life, which also include diarrhoea and weight loss in addition to fatigue and neurological symptoms [72] In some patients, malabsorption would lead to other symptoms consisting of anaemia,

osteoporosis from decreased Vitamin D, osteomalacia due to lack of Vitamin D [110], and general lack of iron, folate or Vitamin B12 that cause anaemia [111]

Since the small intestine can compensate if the effect of CD is limited, up to 38% of patients would remain asymptomatic [87] This would explain the coeliac “iceberg” proposed by one study, which describes the clinically emerging tip of the iceberg that contains only a few diagnosed symptomatic cases, while the larger section of the iceberg that contains asymptomatic and latent cases remains submerged and undetected [112]

1.2.6 Diagnosis and Screening

The diagnosis of CD is often performed by a General Practitioner (GP), using a combination

of clinical examination, and blood tests, and then finally SBB Depending on the level of symptoms, the order of the tests involved would be determined However, all tests must be performed while the patient remains on their normal gluten-containing diet Otherwise, the test results would be false negative and definitive diagnosis compromised

It initially depends on the GP to consider the possibilities of a patient having CD, using differential diagnosis in a variety of clinical presentations [113] In some cases, diagnosis may be delayed due to its varied presentations [102] Besides, a recent meta-analysis in Italy indicated that the ratio of diagnosed to undiagnosed cases of CD was 1:7 [114] Therefore, in order to diagnose a patient for CD, the GP has to pay careful attention to clinical signs including iron deficiency, anaemia or osteoporosis [87]

Serological tests involve autoantibodies testing for anti-reticulin antibodies, anti-gliadin antibodies (AGA), and anti-endomysial antibodies which targets cell surface tissue transglutaminase (tTG), thus it became more specifically known as anti-tissue transglutaminase antibodies (tTGA) [51] Following the advent of increasingly sensitive and specific non-invasive tests, especially in high risk groups, the ability to identify CD in mild and asymptomatic cases is vastly improving [109] The first three histological serology tests are performed based on either indirect immunofluorescence (reticulin, gliadin and endomysium) or ELISA (gliadin and endomysium) [115] The test for IgA anti-tTGA has recently replaced previous tests, by achieving high sensitivity of over 95% and specificity ranging

from 90 to 96% [116] Furthermore, there are selective IgA deficient patients in 3% of CD patients [87], who have IgG that cannot switch to IgA In these patients, IgG anti-tTGA were efficiently used

as an alternative test, with sensitivity and specificity both found to be over 98% [117]

Depending on the clinical diagnostics and serology outcomes, HLA-DQ2 and DQ8 testing might be employed, since it has a high sensitivity in excess of 99% [50] Nevertheless, the specificity

of these genetic tests is not ideal and often gives negative predictive values This is because approximately 30% of the general population, and even higher proportions of high-risk subjects (with associated autoimmune disorders and family history), also carries HLA-DQ2 and DQ8 markers

In any case, SBB remains the internationally approved “gold-standard” for diagnosis [118] Even though the combination of IgA anti-tTGA and total serum IgA are the preferred screening tests, they are not perfect [113] Hence, SBB is required for confirmation of CD diagnosis, as well as for assessment of regression after treatment [119] Most authorities recommend obtaining at least four tissue samples, to avoid false-positive results, as well as improving sensitivity of the test [87] By combining the experience of past studies, one study was able to propose a clinical decision tool that would detect all cases of CD in patients referred for gastroscopy, with 100% sensitivity and 61% specificity [102]

Different screening programs based on antibodies testing have revealed a high prevalence of

CD [101, 112, 120-124] Even though it fulfils the criteria for general population screening as defined

by the World Health Organisation (WHO), mass screening is not yet recommended, because it has not been demonstrated that there is any benefit of screening for asymptomatic individuals [125] The most practical approach would be to screen patients with conditions associated with CD

1.2.7 Current Therapy and Treatment

There is no cure for CD However, the disease can be treated with a strict life-long adherence

to GFD This means that, in order to avoid relapse of inflammatory symptoms and associated complications, gluten must be carefully screened for and completely eliminated from every single meal Hence, in order to achieve absolute compliance to the diet, clinicians must ensure that the patient receive adequate education, motivation, emotional and social support

Before starting the GFD, patients must consult with a dietician to design an intervention strategy to improve the knowledge of CD; especially to know how to distinguish between gluten-containing and gluten-free products [116] It is also important for the patient to fully understand the consequences of untreated CD, as well as the outcome associated with non-compliance to GFD The patient is further encouraged to join a local coeliac society, which would provide crucial guidance for their GDF, as well as receiving emotional and social support opportunities from other members Moreover, the health care team which includes a dietician and a physician has the responsibility to follow-up with the patient’s treatment outcome; to determine the degree of histological improvements,

or whether the condition involves persistent or intermittent symptoms due to known or inadvertent ingestion of gluten [126] Gluten rechallenge with subsequent SBB is no longer required for definitive diagnosis except in unusual cases [87]

Intestinal damage begins to heal within weeks of GFD, and antibody levels eventually decline over months [100] Most of the patients would have experienced a dramatic improvement in symptoms, as the inflammation in their small intestine resolves, and the lining returns to normal However, a minority of patients who have severely damaged intestine won’t improve immediately following GFD These patients require additional medications, such as corticosteroids and other immunosuppressants, to reduce inflammation and accelerate recovery Additional treatment for nutritional deficiency status (e.g iron, folate, Vitamin B12) might also be required for individual cases [87] Furthermore, measurement of bone mineral density to assess for osteoporosis is also recommended

1.2.8 Consequences of untreated coeliac disease

Untreated CD is associated with high morbidity and increased mortality [102] Malignancy has been the most feared complication of CD, with increased risk of small bowel carcinoma, oesophageal and oropharyngeal squamous carcinomas [125, 127] A minority of untreated patients develop a life-threatening condition called thyroid storm, involving uncontrollably high heart rate, blood pressure and body temperature, which could worsen quickly and leads to stroke or heart attack [128] Other life-threatening complications include the development of refractory CD and

enteropathy-associated T cell lymphoma [68], which is a gastrointestinal cancer According to one study, increased risks for many other types of cancers were also found in patients on gluten-containing diets [129] Furthermore, untreated pregnant women with CD have increased risk for preterm birth, having low birth weight babies [130, 131], early menopause, infertility and miscarriage

1.3 RATIONALE

1.3.1 The Link between Graves’ and coeliac disease

Many past studies have reported the association between coeliac and other autoimmune diseases [132-134] More specific studies described the link between coeliac and autoimmune thyroid disease [135-138] But most importantly, the co-existence rate between coeliac and Graves’ disease in individuals with either one of these diseases, was found to be higher compared with the general population [139, 140] One hypothesis for this is that HLA haplotypes B8 and DR3 are more common

in both diseases than the general population [77, 78, 140-142] Other similarities involve both being autoimmune diseases, with multifactorial aetiology, systemic symptoms, and no cure except treatment;

if left untreated, both would contribute to greater risks of failed pregnancy Since all evidences point towards a link between these two diseases, it is logical to propose that the prevalence of CD in GD is higher than the prevalence of CD in a general population, and vice versa

1.3.2 High rate of Vietnamese Graves’ patients at Western Hospital

There has been an increasingly high attendance rate of Vietnamese patients with GD at Western Hospital (WH), as well as the local clinics in the surrounding areas Based on the evidence found in the literature emphasising an important link between Graves’ and coeliac disease in Caucasians, the possibility that Vietnamese patients with GD at WH had an increased risk of CD was considered The incidence of CD in Vietnamese Graves’ patients has not previously been described Therefore, this select population was screened for any signs and serology markers of CD

1.3.3 Early Prevention reduces Morbidity, Mortality and other Risks

Since untreated CD is associated with high morbidity, mortality and an increased risk of complications, it is important to assess whether the risk of CD is increased in Vietnamese GD patients;

in order to provide early preventative treatment against those conditions One study predicted that, as the Asian diet shifts to contain more wheat, rye, barley and oats, more cases of CD will appear [89] This is most applicable in the setting of our project, which focuses on the Vietnamese Community living in Melbourne, Australia

1.3.4 Micronutrient Deficiencies Might Predict coeliac disease

The result of malabsorption, as a complication, suggests that certain micronutrient deficiencies might be established as possible markers for CD If the deficiency of these micronutrients can effectively be used to identify asymptomatic or latent cases of CD, they could perhaps replace IgA anti-tTGA, by providing the advantages of cost-effectiveness and easy accessibility The foundation of this claim comes from suggestions in a previous study that silent CD causes nutritional deficiency of iron, zinc, folate, and the fat-soluble vitamins, D, K, and E [143]

1.3.5 To Establish the Prevalence of coeliac disease in Vietnamese

By examining the link between CD and GD through testing, our pilot study will assess the prevalence of coeliac disease in a Vietnamese population for the first time Previous studies have never specifically focused on an entire population of Vietnamese to screen for CD, or to identify the prevalence of CD in Vietnamese GD patients Therefore, our study will be the first to report the prevalence of CD in a Vietnamese control population, as well as the prevalence of CD in a population

of Vietnamese with pre-existing GD

The prevalence of coeliac disease in Vietnamese participants will also be compared with the prevalence found in Caucasian populations as reported in the literature

2.3 Secondary Aim

The secondary aim of this study is to determine the frequency of micronutrient deficiency in Vietnamese patients with Graves’ disease and in control participants

participants, to allow sufficient power to detect differences in prevalence of IgA anti-tTGA (see section 3.3.1)

3.1.2 Recruitment Strategies

The original plan was to recruit GD patients from the general Endocrinology clinics of WH and RMH, and to recruit control participants from external sources, by utilising various modes of contact and advertisements Over the course of the project, recruitment sources had shifted to include more externally recruited participants, rather than recruits from RMH However, the main source of patients remained at WH

All documents used in the recruitment of participants were fully translated into Vietnamese Most of the time, the English and Vietnamese versions of the same document were used simultaneously for the same purpose The entire list of recruitment strategies employed in our project involves:

Advertisement Poster – More than 100 posters were placed in strategic locations throughout WH

and RMH (see Appendix 7 & 8) English and Vietnamese posters about the project were hung

side-by-side on notice boards, in waiting areas, cafeteria, clinics, next to lifts, walkways, around entrances and parking zone

Endocrinology Clinic – Every Friday morning, the Endocrinology clinic of WH would treat GD

patients Thus, over the recruiting period of our project, the patients’ names and contacts were gathered and compiled into a database, and invitation letters and project flyers were sent home to the patient, to invite them to participate in our study Alternatively, if the patient had come for treatment, they were approached directly A total of 94 patient names were gathered by this method

General Contact – project advertisement was also distributed through family and friends, as well as

neighbours News of the project was either spread by word-of-mouth, invitation letters, project flyers

or phone calls

Groups Contact – many Church groups were contacted, including Footscray CMA Church, Western

New Community Baptist Church, Seddon CMA Church, Melbourne (Yarraville) Church and St Albans CMA Church The Vietnamese organisations that were contacted include the Vietnamese Rheumatism Organisation, Vietnamese Volunteers group, Vietnamese Elderly Organisation, Vietnamese Veteran Organisation and Vietnamese Women Organisation

Invitation Letter – a total of more than 200 letters were directly distributed or sent home to the

potential participants (see Appendix 5 & 6) Out of the 79 letters that were sent to the address of GD

patients gathered from the Endocrinology clinic, only 26 letters generated a response from the receiving patient, which prompted them to accept the invitation and participate in our study Nevertheless, this was considered to be an effective response, and would have been continued if the recruitment requires more participants

Local Clinics – this strategy, as part of the initial recruitment plan, was considered to be unnecessary,

given the success rate of Radio Advertisement However, the strategy had involuntarily provided one participant by referral of one local GP, who had listened to the radio advertisement for our project

Medical Records – the initial plan of recruitment involved reading through the medical records at

WH, in order to identify GD patients for our project However, this plan was not followed, due to the fact that GD patients were readily identifiable from the Endocrinology clinic; plus the number was abounding compared to the number required for our project

Online Invitation – online recruitment was carried out, utilising the Facebook event facility and

profile on Multiply website, both of which allow the promotional information regarding our project to

be displayed to the entire online community In addition, a Promotional Video was made to introduce the potential participant to our research project, as well as reminding them to get in contact for more information before signing-up for participation This video was uploaded onto Facebook, Multiply and YouTube for general access It is currently available at the following addresses (YouTube keywords search “Peter Ebeling”, “Boussioutas”, “Anderson coeliac” or “Khoa coeliac”):

Telephone Invitation – invitation of GD patients by telephone has been trialed as a recruitment

strategy, but was discontinued when its effectiveness was found to be incomparable to sending an invitation letter

Project Flyers – this strategy was modified from the idea of distributing an invitation letter In fact,

the project flyer is a summary of the invitation letter with the main points highlighted, and printed on one half of an A4 page It was designed so that the potential participant would accept the flyer without hesitation, since it has a smaller size and simple content Distribution of the project flyer to different

groups and organisations was relatively easy Thus, over 800 project flyers were dispatched However, the effectiveness of a project flyer cannot be determined since they were also easily discarded

Radio Advertisement – this recruitment strategy was the most effective strategy employed It

generated more than 120 phone calls of interest for participation, contributed to 59 out of 180 participants in the project, and resulted in 31 people being placed on a waiting list, due to the limited number of participants allowed for the project Despite the overwhelming success in recruitment, this strategy was never meant to be one of our main methods of recruitment, due to preformed misconceptions of the cost involved It turned out, as we discovered, that the radio advertisement for research projects which benefit the community can be sponsored free of charge Fortunately, our project was one that screens the general population for latent CD, and in addition provides treatment for any diagnosed cases Hence, we were able to receive free radio advertisement on two different radio stations, as well as an interview with the Vietnamese Language Program, which was broadcasted in the Medical News section entitled “Health is Gold” The first radio station was Stereo

FM 97.4 Vietnamese Radio, where our project advertisement was read once every night from 8 to 10pm (Monday to Friday) and 8 to 12 pm (Sunday), for one whole week The second radio station was SBS Vietnamese Radio 122.4 AM, which broadcasted infrequent advertisements for our project over a two weeks period It was SBS Radio which also hosted the radio interview, conducted entirely

in Vietnamese, to explain in detail about the research of our project The interview broadcast was scheduled between 7:20-7:40pm (section started at 7pm) on Monday 21st July 2008 It is currently available from the following address:

http://www.4shared.com/file/66870654/d4a1ac5b/2008_07_21_-_Vietnamese_Language_Program_On_SBS_Radio_1224_AM_720pm_Interview.html

Short Visit – the shortcoming of this strategy was rapidly realised by the fact that news of the project

spread even slower than General Contact The aim behind this recruitment strategy was to relay advertisement through group leaders and those with wide connection to the general community However, it was determined that the relayed information quickly ceased without any result Therefore, these short promotional visits were stopped soon after a few more than 10 ineffective visiting trips to individual houses

3.1.3 Inclusion Criteria

To be included in the study, every participant had to fulfil all of criteria 1 to 3, plus either one

of criteria 4 or 5:

1 Ability to give informed consent

2 Vietnamese ethnicity → i.e born of at least one Vietnamese parent and have an original Vietnamese surname

5 Graves’ patient → confirmed diagnosis of clinical or subclinical GD based on:

- Biochemical hyperthyroidism prior to treatment (suppressed TSH levels and either high or normal free T4 and T3 levels)

Plus one of the following:

- Graves’ ophthalmopathy, dermopathy or enlarged thyroid gland

- Anti-microsomal antibodies (anti-TPO) or anti-TRAbs

- Thyroid scintigraphy compatible with GD (diffuse increased uptake of tracer)

3 Malabsorptive disorder/Inflammatory bowel disease

4 Renal insufficiency (Creatinine Clearance <60 mL/min)

5 Cirrhosis

1 Signing of the Participant Information and Consent Form (PICF) (see Appendix 1 & 2)

2 Completion of Patient Questionnaires and SF36 Questionnaires (see Appendix 3 & 4)

3 Collection of blood by a research nurse at Pathology of WH

The questionnaires were administered by an experienced member of the research team, who would have given a detailed response to all questions raised about the project

All blood samples were transported after collection to Pathology at RMH for blood tests, including Full Blood Examination (FBE), micronutrient analysis and CD diagnosis

3.2.3 Biochemical Blood Tests

FBE includes haemoglobin (Hb), white cell count (WCC), Platelets (PLT), red cell count (RCC), and mean corpuscular volume (MCV)

Blood test for micronutrient deficiencies:

- Serum folate (sFol)

- Iron (Fe), calcium (Ca) and phosphate levels (PO4)

- 25-Hydroxy-vitamin D (VitD)

Coeliac diagnosis test:

- Total IgA level

- IgA anti-tTGA

3.2.4 Genes Test and Small Bowel Biopsy

All participants with positive results for both total IgA level and IgA anti-tTGA were invited for additional genetic studies and SBB as part of standard clinical care, in order to confirm the diagnosis of CD

3.3 STATISTICAL ANALYSIS

3.3.1 Sample size and Selection

We predicted that a sample size of 50 to 60 in the patient group and 100 to 120 in the control group, would have been able to detect the differences in numbers of individuals with positive anti-tTGA prevalence, based on data in Caucasian with and without Graves’ disease [146] These sample sizes were determined to be sufficient to detect the prevalence of CD in Vietnamese GD patients However, as no previous data exist, this study is exploratory and best considered to be a pilot study

The implication for choosing a total of 150 to 180 participants was based on the fact that one

previous study had recruited at least 100 participants for the sample size [147] This study by Ch’ng et

al 2005, which was performed on Caucasians as the major difference, had almost the same goals and

settings as our study It aimed to investigate the prevalence and evaluate the role of screening for CD

in a GD patient’s cohort, using AGA and anti-tTGA Therefore, having discovered this study in our extensive literature review, we have decided to base our sample size on theirs

3.3.2 SPSS Software Analysis

All data analysis was conducted using SPSS Statistics Version 16 Software The primary parameters (continuous variables) included Age, VitD, CCa, PO4, sFol, Fe, Hb, MCV, IgA and tTGA Secondary parameters (categorical variables) were Sex, VitD status (Severe Deficiency / Moderate

Deficiency / Mild Deficiency / Insufficiency / Replete), Anaemia, Microcytosis (low MCV), Macrocytosis (high MCV), Folate Deficiency, Iron Deficiency, Hypocalcaemia and Hypophosphatemia Separate statistical analyses were carried out for each set of parameters Values

of p < 0.05 were chosen for rejection of the null hypothesis

1 Each one of the primary parameters was assessed for normal distribution of data, using the Kolmogorov-Smirnov normality test For all normally distributed data, the independent or unpaired Student (two-tailed significance) t-test was performed in conjunction with Levene’s test for Equality of Variances Data that were not normally distributed were evaluated using the non-parametric Mann Whitney’s test

2 Each one of the secondary parameters was analysed using the Pearson’s Chi-square (χ2) test,

to determine whether there is any statistical difference between the patients and controls group In situations where the group’s sample size was smaller than 5, Fisher’s exact test were used instead of the χ2 test for statistical comparisons

3.3.3 Limitations

The only perceived limitation was the selection of sample sizes for each of the patients and controls group Unfortunately, we cannot carry out a power calculation nor calculate the sample size based on current available data because none was found to provide the confidence interval or standard deviation (SD) required

3.3.4 Potential Source of Bias

Prior to the project’s commencement, it was determined that the main factors which could affect the prevalence of CD in Vietnamese GD patients were selection bias and geographic separation Since there was no reimbursement involved, selection bias is mainly due to the fact that recruitment for participants comes from the list of patients in the Endocrinology clinic or controls from local clinics, which might not be representative of the whole general population On the other hand, geographic separation suggests different prevalence in different regions, because the tendency for clustering in some regions may possibly be due to disease-modulating environmental triggers [148]

4 RESULTS

4.1 Recruitment of Participants

A total of 180 Vietnamese participants, including 60 Vietnamese GD patients and 120 Vietnamese control participants, were successfully recruited for our project The entire process of recruitment was completed in a record time of one month period, which started from Monday 30thJune and ended on Friday 25th July 2008 By the end of our official recruitment period, 180 signed

PICF, Patient Questionnaires and SF36 Questionnaires (see Appendix 10 & 11), were collected and

stored in a locked filing cabinet as stated in Ethics submission All participant information presented

in this Thesis was listed according to their De-identified code key (see Appendix 9), which was

assigned before participation

4.2 Biochemical Analysis

A total of 180 sets of Biochemical test results (see Appendix 12) were received from

Pathology at RMH and analysed during the course of this project All primary parameters were considered in calculations of the mean, standard deviation, frequency distribution, percentage

frequencies and p-values (see Table I), based on serology results obtained

TABLE I: Baseline Characteristics of the Participants

Primary Parameters Graves' (n=60) Control (n=120) P-values*

IgA anti-tTGA (Units) 6.40±4.71 6.64±5.17 0.757

Data are presented as n (%) or mean ± SD

Coeliac disease was diagnosed in 1 Vietnamese GD patient and 3 Vietnamese control participants, one of whom had borderline IgA anti-tTGA levels (19.9 of 20 Units) These potential CD patients were referred for HLA-DQ tests and SBB as part of standard clinical care By the time this Thesis was completed, the potential CD patients would have had their second blood collection for the

genes test and continued with SBB while waiting for the results, which is currently pending (see Table II) As the result, the prevalence calculated based on anti-tTGA test of CD is 1:60 in Vietnamese

patients with GD, 1:40 in those without GD, and 1:30 in the general population of Vietnamese who

participated in our project (see section 5.2)

TABLE II: Prevalence of coeliac disease by Group

Diagnosis Graves' (n=60) Control (n=120) Prevalence

Careful observation of serology test results had alerted our attention to the fact that there is an existing high prevalence of Vitamin D deficiency in the study population Based on their vitamin D

status, the participants were separated into different groups (see Table III & Figure 1), according the criteria outlined in an Australian and New Zealand Position Statement [149] (see Appendix 10)

TABLE III: Prevalence of Vitamin D Deficiency and Insufficiency by Group

Vitamin D Status Graves' (n=60) Control (n=120) P-values§

Severe Deficiency 0 (0%) 0 (0%) (Not calculated)

§P-values for the difference between controls and Graves’ patients calculated using Chi-square or

Fisher’s exact test

FIGURE 1: Prevalence of Severe, Moderate and Mild Vitamin D Deficiency and Insufficiency

by Group

Note: 88.3% had Vitamin D deficiency or insufficiency, while 55.8% had Vitamin D deficiency alone

In addition to the observed Vitamin D deficiencies, other nutrient deficiencies were studied,

including calcium, phosphate, serum folate and iron (see Table IV and Figure 2)

TABLE IV: Prevalence of other Nutrient Deficiencies by Group

Nutrients Deficiencies Graves' (n=60) Control (n=120) P-values**

Hypocalcaemia (low Ca) 1 (1.7%) 1 (0.8%) 1.000

FIGURE 2: Prevalence of other Nutrient Deficiencies by Group

The prevalence of anaemia (low Hb) were also considered, while examined together with the

MCV (microcytosis and macrocytosis) (see Table V & Figure 3)

TABLE V: Prevalence of Anaemia, Microcytosis and Macrocytosis by Group

Parameters Graves' (n=60) Control (n=120) P-values**

FIGURE 3: Prevalence of Anaemia, Microcytosis and Macrocytosis by Group

5 DISCUSSION

5.1 Baseline Characteristics

The only baseline characteristics that were different at baseline between groups were sex and folate

It was not practically possible to match participant for sex, due to initial recruitment and time

limitations of the project (see section 5.6) As the result of recruitment, the ratio of male to female in

the control group was observed to be more than twice that of the GD patients group This would be

expected, given that the incidence of GD is five to ten folds in women compared to men (see section 1.1.2), which explains the low number of men recruited for the GD patient group In contrast, a higher

number of men in the control group cannot be avoided, since the initial recruitment in this group preceded that of the patients group This was due to the difference in time it took for a response to be received from GD patients compared with control participants, as well as the response rate for male patients, both of which could not be anticipated before the study

There was a statistically significant difference in mean sFol levels between the control and Graves’ disease group, which was due to 10 indeterminate cases (between normal and deficient levels) and 1 deficient case in GD patients, as compared with 2 indeterminate cases in control group Based

on the gathered data, we speculated that the incidence of folate deficiency, and possibly anaemia is

higher in Vietnamese GD patients than controls (see section 5.4) However, after conducting an

extensive search of the current literature for an association between folate deficiency and GD, we did not find any supporting evidence for this hypothesis

5.2 Prevalence of coeliac disease in Vietnamese Graves’ disease Patients

IgA anti-tTGA test results were found to be negative for the correlation between CD and GD

in Vietnamese participants (see Primary Aim 2.1) This is because the prevalence of 1:60 calculated

for CD in Vietnamese GD patients is approximately equal or less than that of the Vietnamese control

group (see Secondary Aim 2.2), which is 1:60 (or 1:40 if the participant with borderline IgA

anti-tTGA is included) As the result, the hypothesis of our pilot study was disproved, since we had determined that the risk of CD appears to be similar in Vietnamese with or without GD Nevertheless,