On diet in ankylosing spondylitis

Björn Sundström

Department of Public health and clinical medicineDivision of Rheumatology Umeå University, SwedenUmeå 2011

On diet in ankylosing spondylitis

Björn Sundström

Institution för folkhälsa och klinisk medicinUmeå universitetUmeå 2011

Responsible publisher under swedish law: the Dean of the Medical FacultyThis work is protected by the Swedish Copyright Legislation (Act 1960:729)New Series No 1440, ISSN 0346-6612ISBN 978-91-7459-272-6Elektronisk version tillgänglig på http://umu.diva-portal.org/Printed by: Print & mediaUmeå, Sweden 2011

“Believe those who are seeking the truth. Doubt those who find it.”

André Gide (1869 - 1951)

i

TABLE OF CONTENTS

ABSTRACT …………………………………………………………………….……………………...iii

ENKEL SAMMANFATTNING PÅ SVENSKA ……………………..………………....v

ABBREVIATIONS ………………………………………………………………………….……..vii

BACKGROUND

Ankylosing Spondylitis .…………………………………………………………………….…………1

Comorbidity in ankylosing spondylitis……………………………………………………….…3

Clinical assessment of ankylosing spondylitis .……………………………………………...4

Assessment of dietary intake .……………………………………………………………………...5

Diet in ankylosing spondylitis and associated diseases .…………………………..……..7

Diet in ankylosing spondylitis ..………………………………………………………......……..7

Diet in psoriatic arthritis ...…………………………………………………………………………8

Diet in inflammatory bowel disease...………………………………………………………….9

Rationale for studying diet in ankylosing spondylitis……………………………………10

AIMs

General aim.………………………………………………………………………………….………….13

Specific aims .……………………………………………………………………………..…………….13

METHODS

Subjects and study design

Paper I………………………………………………………………………………………...14

Paper II…………………………………………………………………...…………………..15

Paper III…………………………………………………………….………………………..15

Paper IV…………………………………….………………………………………………..15

Assessments

Assessments of disease……………………………………………………………..…..16

Assessments of dietary intake………………………………………………………..17

Background factors and other specified assessments……………………...18

Laboratory performance and routines ..………………………………...……..20

Ethical aspects ………………………………………………………………………………….……..20

Statistical methods…………………………………………………………………………….……..20

ii

RESULTS AND DISCUSSIONS

Paper I ……………………………………………………………………………………….…….……..22

Paper II …………………………………………………………………………………...……….……..25

Paper III ………………………………………………………………………………….…….………..27

Paper IV………………………………………………………………………………….……..………..29

General discussion, limitations and future perspectives…………….………..………..31

CONCLUSIONS ……………………………………..……………………….………..35

ACKNOWLEDGEMENTS ………………..………………………………………..36

REFERENCES…………………………….…………………………………….….....37

iii

ABSTRACTThe aim of this thesis was to examine the role of diet in ankylosing spondylitis

(AS). Patients were examined in: i) a postal questionnaire survey of dietary

habits and gastrointestinal (GI) symptoms; ii) a study on biomarkers of diet and

disease activity; iii) a comparison of cardiovascular risk factors with the general

population using data from the Västerbotten Intervention Programme (VIP),

and; iv) a 21-week omega-3 fatty acid supplementation study regarding the

effects on disease activity.

The postal survey (111 respondents) revealed no correlation between dietary

habits and disease activity measured by the Bath Ankylosing Spondylitis Disease

Activity Index (BASDAI). However, GI problems, and in particular GI pain,

were prevalent in patients with AS irrespective of NSAID usage.

Gastrointestinal pain was predicted by higher BASDAI and a higher

consumption of vegetables. Overall, 30 (27%) of the patients experienced an

aggravation of gastric symptoms when consuming certain foods. In the study of

biomarkers (n=66) no correlation was found between diet and disease activity

as assessed by BASDAI. There were, however, positive correlations between

BASDAI and the content of arachidonic acid (AA) in plasma phospholipids

(rs=0.39, p<0.01) and the estimated activity of the enzyme delta-5-desaturase

(rs=0.37, p<0.01). This may reflect a process involved in the inflammation

associated with AS that requires further investigation. Comparing data from the

VIP for patients (n=89) and controls showed no significant differences

regarding diet, physical activity or smoking. Nonetheless, more pronounced

correlations between blood lipids and diet were identified among patients than

in controls. Furthermore, the levels of cholesterol and triglycerides were lower

in patients compared with controls. Lastly, in the supplementation study, a

high-dose of long-chain omega-3 fatty acids (4.55 grams/day) was found to

lower disease activity, as measured by BASDAI, whereas low-dose treatment

(1.95 grams/day) caused no change.

In conclusion, within a group of Swedish AS patients we found no correlation

between ordinary dietary habits and disease activity. Diet in western

iv

populations of patients with AS may, however, be of importance for gastric

symptoms and for cardiovascular risk factors. The finding of a lowered disease

activity in patients on high-dose supplementation with long-chain omega-3 fatty

acids indicates that a radical dietary shift may influence disease activity. The

findings of a positive correlation between disease activity and plasma AA, and

the decreased levels of blood lipids imply the need for further studies into fatty

acid metabolism in AS.

v

ENKEL SAMMANFATTNING PÅ SVENSKA

Ankyloserande spondylit (AS) är en kronisk reumatisk sjukdom som tillhör

gruppen spondylartropatier. Sjukdomen drabbar i huvudsak stora leder och

ryggrad, där ryggraden i svårare fall kan bli helt förbenad. Det är vanligt att

patienter med AS har inflammatoriska förändringar i tarmslemhinnan som

påminner om de som ses vid den inflammatoriska tarmsjukdomen Morbus

Crohn. Vidare har man sett en ökad dödlighet i hjärt-kärl sjukdomar bland

patienter med AS. Sjukdomens förekomst har visats överensstämma väl med

förekomsten av den genetiska markören HLA-B27. Både sjukdomen och

förekomsten av HLA-B27 är vanligare ju längre norrut man kommer. Allra

vanligast är HLA-B27 hos vissa grupper av eskimåer. Trots det har man i

studier av eskimåbefolkningar sett att förekomsten av AS är lägre än vad man

förväntade sig. I andra studier har det föreslagits att ett högt intag av omega-3

fetter bland eskimåbefolkningar skulle kunna leda till mindre förekomst av

inflammatoriska sjukdomar som till exempel reumatoid artrit och psoriasis.

Trots att både förekomsten av förändringar på tarm och den låga förekomsten

av AS hos eskimåer är beskriven sedan tidigare, finns det förvånansvärt få

studier som har undersökt hur kosten påverkar sjukdomsaktiviteten vid AS.

Syftet med denna avhandling var därför att undersöka kostens effekter vid AS.

Avhandlingen baseras på två kohorter av patienter. En kohort med

ursprungligen 166 patienter som svarade på en enkätundersökning, deltog i en

studie av biomarkörer för kost och sjukdomsaktivitet och blev undersökta med

en jämförelse av data från Västerbottensprojektet (VIP). Den andra kohorten,

som bestod av 24 patienter, provade ett kosttillskott av omega-3 fettsyror under

21 veckor. I enkätundersökningen, där resultaten baseras på svaren från 111

patienter, kunde man inte se något klart samband mellan kost och

sjukdomsaktivitet hos patienterna. Däremot såg man att magbesvär, främst

magsmärtor, var vanliga oberoende av om man använde sig av icke-steroida

inflammationsdämpande läkemedel, så kallade NSAIDs, som ofta ger

magbesvär. Magsmärtor var vanligare bland patienter med högre

sjukdomsaktivitet, och bland de som konsumerade mer grönsaker. Mer än var

vi

fjärde patient med AS upplevde att de fick mer magbesvär när de åt vissa

livsmedel som grönsaker/frukt, mjölkprodukter, fet mat och mjölrik mat.

Upplevelsen att bli sämre av dessa livsmedel påminner om vad patienter med

inflammatorisk tarmsjukdom har beskrivit. Inte heller i biomarkörstudien, där

66 patienter deltog, kunde några tydliga samband ses mellan kost och

sjukdomsaktivitet. Däremot sågs samband mellan sjukdomsaktivitet och

halterna av arakidonsyra i fosfolipider, vilket kan spegla processer involverade i

inflammation vid AS. I jämförelsen med registerdata där 89 patienter

jämfördes med 356 personer från övriga befolkningen i Västerbotten kunde inte

några signifikanta skillnader när det gäller kost, fysisk aktivitet och rökning ses.

Det sågs däremot mer tydliga samband mellan blodfetter skadliga för hjärta och

kärl, och kost bland patienterna jämfört med kontrollerna. Patienterna

uppvisade också signifikant lägre nivåer av kolesterol och triglycerider jämfört

med vad som sågs bland övrig befolkning, vilket kan göra att man underskattar

risken för framtida hjärt-kärlsjukdom hos patienter med AS. Slutligen, i studien

där patienterna fick prova ett kosttillskott med omega-3 fetter sågs att de nio

patienter som fullföljde studien med den högre dosen av 4,55 gram/dag

fettsyror fick sänkt sjukdomsaktivitet. Detta sågs inte hos de nio patienter som

fick den lägre dosen 1,95 gram/dag.

Sammanfattningsvis fann vi inte något samband mellan kost och

sjukdomsaktivitet hos svenska patienter med AS. Den vanliga kosten som de

undersökta patienterna konsumerade kan däremot vara av betydelse för

magbesvär och för hjärt-kärlsjukdom. De positiva effekterna på

sjukdomsaktivitet av hög dos omega-3 som kosttillskott pekar på att en radikal

förändring i kost skulle kunna påverka sjukdomens aktivitet. Sambandet

mellan sjukdomsaktivitet och nivåer av arakidonsyra i fosfolipider, samt de

lägre nivåerna av blodfetter i patientgruppen motiverar ytterligare studier av

fettmetabolismen hos patienter med AS.

vii

ABBREVIATIONS

AA arachidonic acid

ALA alpha-linolenic acid

ALT alanine aminotransferase

AS ankylosing spondylitis

AST aspartate aminotransferase

ASDAS ankylosing spondylitis disease activity score

BASDAI Bath ankylosing spondylitis disease activity index

BASFI Bath ankylosing functional index

BMI body mass index

CBC complete blood count

CRP c-reactive protein

CVD cardiovascular diseases

D5D delta-5-desaturase

D6D delta-6-desaturase

DHA docosahexaenoic acid

DHI diet history interviews

DPA docosapentaenoic acid

EPA eicosapentaenoic acid

ESR erythrocyte sedimentation rate

FADS-1 fatty acids desaturase gene 1

FADS-2 fatty acids desaturase gene 2

FFQ food frequency questionnaires

FIL food intake level

FODMAPs fermentable oligo-, di-, and mono-saccharides and polyols

GI gastrointestinal

GT gammatransferase

Hb haemoglobin

HDL high density lipoprotein

HLA-B27 human leukocyte antigen B-27

hs-CRP high sensitive c-reactive protein

IBD inflammatory bowel disease

viii

IFN interferon

LA linoleic acid

kJ kilojoule

MCP monocyte chemotactic protein

MMP-3 matrix metalloproteinase 3

MUFA monounsaturated fatty acid

NSAID non steroidal anti-inflammatory drug

PAL physical activity level

PsA psoriatic arthritis

PUFA polyunsaturated fatty acid

RA rheumatoid arthritis

ReA reactive arthritis

REE resting energy expenditure

SCD stearoyl-CoA desaturase

SCFA short-chain fatty acid

SFA saturated fatty acid

SpA spondyloarthropathies

TG triglycerides

TNF-α tumour necrosis factor alpha

uSpA undifferentiated spondyloarthropathy

VAS visual analog scale

VIP Västerbotten intervention programme

ix

List of original papers

This thesis is based on the following papers:

I. Sundström B, Wållberg-Jonsson S, Johansson G. Diet, disease activityand gastrointestinal symptoms in ankylosing spondylitis. ClinRheumatol 2011: 30:71-6

II. Sundström B, Johansson G, Cederholm T, Kokkonen H, Wållberg-Jonsson S. Plasma phospholipids are related to disease activity inankylosing spondylitis. J Rheumatol 2011-0575.R2 [in press]

III. Sundström B, Johansson G, Johansson I, Wållberg-Jonsson S. Diet andcardiovascular risk factors among patients with ankylosing spondylitisin comparison with the general population. [Manuscript]

IV. Sundström B, Stålnacke K, Hagfors L, Johansson G. Supplementation of Omega-3 fatty acids in patients with ankylosing spondylitis. Scand J Rheum 2006; 35:359-362

Paper I reprinted with kind permission from Springer-Verlag. Paper II reprinted with kind permission from The Journal of Rheumatology Publishing Company. Permission to reprint paper IV in this thesis was not necessary.

x

Prologue

"Despite our efforts to locate and examine as many cases as possible, we still

did not find AS to be a common disorder in these populations. This finding may

be considered surprising because of the known high frequencies of HLA-B27 in

the populations. Many of the cases we did identify were mild and self-limited."

-Georgiana Boyer, on the prevalence of spondyloarthropathies in Alaskan Eskimos.

Journal of Rheumatology 1994.

Twelve years ago, while searching for articles regarding physical activity in

patients with ankylosing spondylitis I stumbled on Boyer’s articles on the

prevalence of spondyloarthropathies among Inuits and other circumpolar

populations. Although, Georgiana Boyer mentioned both physical activity and

the sterile environment around the polar circles as possible explanations to the

low prevalence and mild cases found, I believed that there was a significant

question missing in the discussion. The diet, which is so unique for these

populations, was not discussed. The importance of dietary fat quality, a topic

that had gained considerable attention among other inflammatory diseases, was

left open. Reading that article, 12 years ago, I could never have imagined that

this issue would become the subject of a dissertation for me.

Umeå, October 2011

xi

1

BACKGROUND

Ankylosing SpondylitisAnkylosing spondylitis (AS) is a chronic inflammatory disease that belongs to the group of spondyloarthropathies (SpA) together with psoriatic arthritis (PsA), reactive arthritis (ReA), arthritis associated with inflammatory bowel disease (IBD) and undifferentiated spondyloarthropathy (uSpA)(Sieper, 2006). The disease is clearly distinguished immunologically and pathologically from the most well-known rheumatic inflammatory disease, rheumatoid arthritis (RA) (Inman, 2009; Lories, 2009). Although, the disease had been described by several authors in the latter part of the 19th century, it was not until the first diagnostic criteria, i.e., the Rome criteria proposed in 1961 (Hellgren, 1963), and the beginning of the application of the rheumatoid factor test in RA (Rose, 1948), that the disease clearly started to be treated and investigated as an independent disease and not just as a variant of RA. The diagnostic criteria for AS have evolved over the years and during the last two decades the modified New York criteria have been used (Figure 1)(van der Linden, 1984). There is currently a transition to new diagnostic criteria and new naming schemes in SpA, according to which, a distinction is to be made between axial and peripheral disease (Rudwaleit, 2010).

A. Clinical criteria i. Low back pain and stiffness which improves with exercise, but is not relieved by rest ii. Limitation of motion of lumbar spine in both the sagittal and frontal planes iii. Limitation of chest expansion relative to normal values corrected for age and sex

B. Radiological criteria Sacroiliitis grade 2 bilaterally, or sacroiliitis grade 3-4 unilaterally

Grading Definite ankylosing spondylitis is diagnosed if the radiological criterion is associated with at least one clinical criterion Probably ankylosing spondylitis is diagnosed if either: i. Three clinical criteria are present ii. The radiological criterion is present without any signs or symptoms satisfying the clinical criteria

Figure 1 The modified New York criteria for ankylosing spondylitis (van der Linden 1984).

2

One characteristic of the disease is enthesitis of ligaments and tendon

insertions, which may lead to tissue calcification and bone ankylosis (Lories,

2009). Classically, AS involves the sacroiliac joints and the anterior

longitudinal ligament of the spine, of which the former is an important part of

the diagnostic criteria.

Treatment options for the disease have previously been limited, and largely

consisted of physical exercise and administration of non steroidal anti-

inflammatory drugs (NSAIDs). In recent years, biological treatments have been

introduced, i.e., inhibitors of tumour necrosis factor alpha (TNF-α) (Braun,

2003). Although the new biological pharmaceuticals have high efficacy, not all

patients are eligible for them. According to current Swedish guidelines, patients

with AS may be treated with biological pharmaceuticals if they exhibit a high

disease activity score and conventional treatment has failed.

The prevalence of AS is different in various regions of the world and it has been

shown to correspond well with the presence of the genetic marker HLA-B27

(Gran, 1993). For Sweden, a general prevalence of 0.5 % was previously

suggested (Engström-Laurent, 1994), but recent studies point to a lower rate of

about 0.12 % (Haglund, 2011). The prevalence of HLA-B27 is higher in the

northern parts of the world (Bjelle, 1982), with a particularly high frequency

among the Inuits in arctic populations (Khan, 1996). Despite this high

prevalence of HLA-B27 (25-40%) an unexpectedly low incidence of AS has been

revealed among Alaskan Inuits (Boyer, 1988; Boyer, 1994; Lawrence, 1996).

Many of the cases diagnosed in these studies were of mild nature, which has

raised questions as to whether lifestyle or environmental factors may influence

the onset and future progress of the disease. Previous studies on Inuit

populations have shown that a high intake of polyunsaturated fats of omega-3

series fatty acids probably result in a lower incidence of inflammatory diseases

such as RA and psoriasis (Horrobin, 1987; Kromann, 1980).

3

Comorbidity in ankylosing spondylitis

Patients with AS have been shown to have an increased prevalence of

cardiovascular diseases (CVD) (Han, 2006). In addition to ischaemic heart

disease (Hollan, 2008), the cardiac manifestations also include conduction

disease, aortic root disease and valve disease (Bergfeldt, 1982; Momeni, 2011).

Cardiovascular diseases are a major cause to the higher mortality among

patients with AS (Zochling, 2009). In addition to an accelerated coronary

atherosclerotic process, the increased mortality may also be due to the

conduction disorders (Dik, 2010). The inflammation and fibrosis seen in the

aortic root and mitral valve among AS patients (Roldan, 1998), may, if extended

to the interventricular septum, the atrioventricular node, and the bundle

branches, play an important role in the conduction disturbances. Interestingly,

it has been suggested, but unfortunately not elucidated further, that the genetic

marker HLA-B27, even without the presence of inflammatory diseases such as

AS, is a predictor of conduction disorders (Bergfeldt, 1988; Bergfeldt, 1997).

In the context of CVD, it may be noted that the NSAIDs, which form the basic

treatment for AS, may also contribute to an increased rate of CVD in the general

population (McGettigan, 2006). Although the cardiovascular risks of NSAID

treatment in AS have been discussed (van der Linden, 2008), no studies have

been performed aiming to evaluate which influence NSAIDs might have on the

incidence of CVD among patients with AS.

Studies have shown that there are linkages between IBD and SpA (Mielants,

1991; Mielants, 1996). Many patients with AS have inflammatory changes in

the intestinal mucosa almost identical in nature to those that patients with

Crohn's disease exhibit at an early stage of disease progression (De Vos, 1989;

Mielants, 1991). Furthermore, seven percent of patients with AS develop

Crohn’s disease later in the disease course (De Vos, 1996).

Patients with AS are also affected by osteoporosis to a higher degree (Will, 1989;

Han, 2006), although it has been suggested that advanced stages of AS may

have increased bone mineral density as assessed by dual-energy X-ray

absorptiometry (Mullaji, 1994). High disease activity has been shown to

4

increase the risk of amyloidosis (Gratacos, 1997; Singh, 2007). Nor is

pulmonary involvement uncommon among patients with higher disease activity.

This was previously suggested to be a result of the stiff chest and reduced lung

function that are consistent with a more severe disease (Appelrouth, 1975;

Feltelius, 1986). However, more recent studies point to a parenchymal

involvement of the lungs (Kiris, 2003; Senocak, 2003)

Clinical assessment of ankylosing spondylitis

Clinical assessment of AS has traditionally been hampered by the absence of

valid laboratory biomarkers. Traditional tests for inflammatory diseases, such

as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR) and

orosomucoid, are of no or little value (Sheehan, 1986; Spoorenberg, 1999). The

assessment of disease activity and functional capacity has, therefore, been

directed to patient self-administrated rating scales. The Bath ankylosing

spondylitis disease activity index (BASDAI) and Bath ankylosing functional

index (BASFI) have been the gold standards since their introduction in 1994

(Calin, 1994; Garrett, 1994). The Bath questionnaires have been translated to

several languages with the Swedish versions being available since 1999

(Cronstedt, 1999; Waldner, 1999). The BASDAI evaluates disease activity as a

calculated score based on six visual analog scales (VAS) regarding symptoms

during the previous week. The mean of quantity and quality of morning

stiffness added to values for fatigue, spinal pain, peripheral joint pain or

swelling and tender areas represents the final BASDAI score. The BASFI

questionnaire comprises eight VAS regarding the patients’ ability to perform

activities associated with daily life: putting on socks, bending forward, reaching

up, getting out of a chair, getting up from the floor, standing unsupported,

climbing steps, and lastly, looking over the shoulder. The BASFI also contains

two additional questions regarding the ability to deal with demanding activities

and to handle a full day of activities. The mean of these ten items represents the

final BASFI score.

In recent years, a composite instrument for assessing disease activity has

evolved, i.e., the ankylosing spondylitis disease activity score (ASDAS) (van der

Heijde, 2008; Boers, 2009; Lukas, 2009). The ASDAS combines 5 disease

5

activity variables: CRP and four self-reported items assessed using VAS. The

four self-reported items are: back pain, duration of morning stiffness, peripheral

pain/swelling, and patient global assessment of disease activity. In addition,

there is an alternative version using ESR instead of CRP. The latest work on the

ASDAS instrument, including cut-off values and nomenclature for the disease,

was presented at the same time as the writing of this thesis (Machado, 2011).

In the recent years, work has also been undertaken to identify biomarkers of

radiological progression in an attempt to identify patients with rapid disease

progression. A proposed predictor of radiological progression in patients with

AS is serum matrix metalloproteinase 3 (MMP-3) (Maksymowych, 2007).

MMP-3 is also significantly increased in patients with active AS but does not

correlate significantly with conventional variables used to assess disease

activity. Serum levels of MMP-3 are lowered, together with ESR, CRP and

BASDAI, during treatment with TNF-α inhibitors (Wendling, 2008).

Assessment of dietary intake

Methods for assessment of food intake are dietary recalls, food records, diet

history interviews (DHI) and food frequency questionnaires (FFQ). Dietary

recall, is a retrospective interview method whereby a person's dietary intake is

estimated over a previous, shorter period, usually 24 hours (Wiehl, 1942). Food

records is a prospective method in which participants are asked to compile a

diary-like list of foods and in what quantity they consume these foods for a

specified period, usually three to seven days (Barrett-Connor, 1991).

Estimation of portion sizes is usually performed by weighing, but other

documentation routines, such as photographs, can also be used. The diet

history interview explores food and meal patterns, consumption frequencies and

portions sizes, i.e., the participants are asked what they usually eat, in contrast

to what they actually have eaten (Burke, 1947).

A widely used method to establish dietary intake is the FFQ (Barrett-Connor,

1991). The investigation is usually performed by self-administration, but can

also be performed through an interview. An FFQ comprises a preformed list of

food items and their appropriate frequencies for consumption. It often contains

6

additional questions regarding portion sizes, to form a semi-quantitative food

frequency questionnaire, which can form the basis for a nutritional content

calculation. The main advantage of this method is that it is fast and efficient,

and that it requires few resources for either the participants or for the

investigators, making it suitable for larger epidemiological surveys. The major

disadvantage is that it only records those foods stated in the questionnaire. It

needs, therefore, to be carefully designed and validated to incorporate those

foods consumed in the region, or in the culture, of the particular investigation.

Since food habits also change over time in the community, there is a need for

the FFQ to be up-to-date. Due to its nature of recording only the consumption

of the stated food stuffs, there is a risk of under-estimating food intake in a short

questionnaire and over-estimating in long questionnaires.

To complement dietary surveys, biomarkers have been used during the past few

decades. For instace blood samples or adipose tissue can be sampled and

analysed to reflect the composition of fat intake over a shorter and/or longer

time (Arab, 2002; Hodson, 2008). When analysing samples for fatty acid

composition, it should be noted that samples of different origin may contain

slightly different proportions of fatty acids; furthermore, their different turnover

rates will reflect diet in different time periods. For example, analysing

erythrocyte membranes will reflect dietary intake over a longer time period

compared with plasma phospholipids - due to the longer turnover rate of

erythrocytes. When analysing adipose tissue samples, the origin of the sample

may influence the results, e.g., a sample from the waist may contain different

proportions of saturated fatty acids (SFA) and mono-unsaturated fatty acids

(MUFA) than a sample from gluteal subcutaneous adipose tissue, although the

content of polyunsaturated fatty acids (PUFA) seems to be approximately the

same (Malcom, 1989). A major problem with measurement of dietary

biomarkers is that they are relatively costly, and they require a considerable

amount of time and effort for both investigators and participants, e.g., blood

and tissue sampling or collection of urine or faeces. Therefore biomarkers are

usually used with restraint during dietary surveys and selected with care for

specific aims.

7

Diet in ankylosing spondylitis and associated diseases

Diet in ankylosing spondylitis

Although patients with AS have reported an aggravation of disease activity with

certain foods (Haugen, 1991) and there seems to be a relationship between AS

and the gut, surprisingly few studies have been published regarding AS and diet.

At the time of the planning these studies in 2004 a search on PubMed using the

keywords “Ankylosing” and “Diet” revealed only 32 hits. With the exclusion of

articles not written in English and articles investigating the presence of coeliac

disease in AS, only three articles remained which discussed diet in a clinical

manner. Two articles discussed the role of the Klebsiella bacteria in the diet.

The first article, an intervention study by Shinebaum et al. examined whether it

was possible to lower the disease activity by dietary elimination of possible

sources of Klebsiella, such as cold meat and salads (Shinebaum, 1984). The

study failed to show any differences between the control group and

experimental group. The second article, by Ebringer and Wilson, discussed

experiences of a diet aimed at reducing the substrate for the Klebsiella species in

the colon by reducing the dietary intake of starch (Ebringer, 1996). However, to

date, no controlled study has been presented in which a low-starch diet is

included in the treatment of patients with AS. Lastly, in a letter, Appelboom, et

al. describe a small study on the exclusion of milk products among patients with

AS, SpA and patients with RA (Appelboom, 1994). Patients with AS or SpA

reported good efficacy of the diet, while none of the patients with RA responded

to the dairy product exclusion.

The role of vitamin D in AS has been discussed by several authors, although not

usually as a dietary factor affecting disease activity. The primary topics of these

studies have been the bone transformation and the osteoporosis present in AS

patients (Lange, 2001; Lange, 2005; Mermerci, 2010). These studies suggest

that serum levels of vitamin D are lower among AS patients than among

controls. The reason(s) for this is unknown, but both bacteria-induced vitamin

D receptor dysfunction (Waterhouse, 2009) and associations with levels of

human transglutaminase (Teichmann, 2009) have been discussed.

8

The influence of coeliac disease in rheumatic diseases has also been discussed

by several authors (Adelizzi, 1982; Bourne, 1985). In 1985, 160 patients with

coeliac disease were screened for AS, with not a single case being identified

(Bourne, 1985). Later Kallikorm et al., screened 74 patients with SpA and found

9 patients with elevated anti-gliadin antibodies. In one patient, anti-endomysial

antibodies were also found and a small bowel biopsy confirmed the diagnosis of

coeliac disease in this patient (Kallikorm, 2000). In 2009, Togrol et al.,

published a case-control study in which 30 AS patients and 19 controls were

screened for coeliac disease (Togrol, 2009). Eleven AS patients had elevated

levels of anti-gliadin antibodies and three also had raised levels of anti-

endomysium antibodies. One patient was later diagnosed with coeliac disease

after biopsy. In contrast to Togrol’s and Kallikorm’s studies, Riente et al.

concluded that antibodies indicating coeliac disease are not elevated among

patients with AS, nor those with PsA (Riente, 2004). Lastly, Tiechman et al.

found higher levels of transglutaminase among AS and PsA patients correlated

inversely with serums levels of vitamin D (Teichmann, 2009).

Diet in psoriatic arthritis

Scientific knowledge of dietary influences in patients with PsA is also limited.

Anti-inflammatory effects of omega-3 fatty acids have been shown in two

studies (Lassus, 1990; Madland, 2006). In 1990, Lassus et al., presented the

results from an open study on 80 Finnish patients with psoriatic disease, of

whom 34 had PsA. After 8 weeks with a daily supplementation of 1.9 grams

long-chain omega-3 fatty acids the patients with PsA exhibited significant

decrease in their joint pain (Lassus, 1990). In 2006, Madland, et al. presented

the effect(s) of two weeks supplementation with seal oil containing a daily dose

of 6.1 grams long-chain omega-3 fatty acids (Madland, 2006). The 20 patients

studied exhibited a modest improvement in the global assessment of their

disease and a trend towards a decrease in tender joints count compared with the

20 controls receiving soy oil.

Supplemental oral vitamin D has been investigated in two small studies

(Huckins, 1990; Gaal, 2009). In an open study, 10 patients were treated with

oral 1,25-dihydroxyvitamin D3. Seven patients improved in terms of tender

9

joint count; however, two patients were unable to receive therapeutic doses due

to hypercalciuria (Huckins, 1990). A more recent study investigated the effect

of oral alphacalcidol on 10 patients with PsA with an additional 9 acting as

controls. The study concluded that oral alphacalcidol had effects on

immunological parameters and on the DAS-28 score (Gaal, 2009). As in AS,

the presence of coeliac disease and its associated antibodies has been

investigated in PsA, partly in the same studies. Elevated levels of IgA anti-

gliadin antibodies compared with controls have been described. Patients with

elevated levels also had higher disease activity compared with other patients

(Lindqvist, 2002), but conflicting results have been reported (Kia, 2007).

Diet in inflammatory bowel disease

Only a few dietary studies have been specifically aimed at IBD-associated

arthritis and arthralghia. In two studies, nasoduodenally administrated seal oil,

rich in long-chained omega-3 fatty acids, was shown to improve joint pain in

IBD (Bjorkkjaer, 2004; Bjorkkjaer, 2006). In a later pilot study using oral seal

oil and cod liver oil only trends in the reduction of joint pain could be noted,

although there were significant changes on fatty acid composition in blood

samples (Brunborg, 2008). One small open study regarding the use of

probiotics against arthralgia in IBD patients points out a positive effect on

Ritchies' articular index, but due to the study design and a drop-out rate of

almost half of the study group, it is difficult to draw conclusions (Karimi, 2005).

The use of probiotics has also recently been evaluated in AS patients, revealing

no significant clinical effect although the patients in the intervention group did

exhibit numerically lower ratings on the BASDAI indices (Jenks, 2010).

Studies have suggested that dietary fibres exhibit an anti-inflammatory effect in

the colonic lumen of patients with IBD by inducing the production of short-

chain fatty acids (SCFA). Butyrate has predominantly been suggested to act as a

substrate for colonic epithelial cells accelerating the intestinal repairing process

and thereby preserving intestinal integrity (Torres, 2008). However, there

seem to be cases in which the intake of dietary fibres is not beneficial; for

example, dietary fructo-oligosaccharides appear to decrease resistance to gut

bacteria in mice fed a diet low in calcium (Ten Bruggencate, 2004; Guarner,

10

2007). Gibson and Shepherd (Gibson, 2005) also challenged the current view

on fibres when they presented their hypothesis on fermentable oligo-, di-, and

mono-saccharides and polyols (FODMAPs). They suggested that poorly

absorbed, but highly fermentable carbohydrates could lead to high

concentrations of SCFA and lactic acid, which lower lumen barrier tolerance and

may have an osmotic effect in the bowel.

Because of the severe gastrointestinal problems experienced in IBD, many

patients avoid certain kinds of foods for fear of initiating a relapse of their

symptoms. Jowett, et al. found that patients most commonly avoided milk, fruit

and vegetables, however this approach had little or no effect on relapse rate

(Jowett, 2004). There are no greater differences between ulcerous colitis and

Crohn’s disease as to which foodstuff the patients choose to avoid (Ballegaard,

1997).

Rationale for studying diet in ankylosing spondylitis

Omega-3 fatty acids may have an anti-inflammatory effect in PsA and IBD-

associated arthritis (Lassus, 1990; Bjorkkjaer, 2004; Bjorkkjaer, 2006;

Madland, 2006). The omega-3 fatty acids, as well as the omega-6 fatty acids,

are essential for man due to the inability of mammals to insert a double bound

at the third or sixth carbon position from the methyl end (CH3) of a fatty acid.

The methyl end of a fatty acid is also referred to as the “omega end” or “n end”.

The unsaturated fatty acids were originally named for where the first double

bound occurred, e.g., “omega minus 3” or “n minus 3” fatty acids, thus giving

rise to the naming scheme that is commonly used today. Although both

omega-3 and omega-6 fatty acids are classified as essential, man has a limited

ability to endogenously produce longer omega-3 and -6 fatty acids from the

basis of shorter omega-3 and -6 fatty acids by a series of elongation and

desaturation steps (Figure 2). Therefore, only the shorter alpha-linolenic acid

(ALA) and linoleic acid (LA) are strictly essential. Due to the inability of

mammals to convert between omega-3 and -6, the endogenous production of

longer omega-3 and -6 fatty acids are dependent on the proportions between,

and the availability of, the shorter ALA and LA. The diet of industrialised

western countries are generally abundant in omega-6 fatty acids, most

11

commonly by the omega-6 fatty acid LA which is present in vegetable oil such as

sunflower, cottonseed and soy oils. The main sources of omega-3 fatty acids in

the western diet are the vegetable canola and flaxseed oils. The longer omega-3

and -6 fatty acids, such as arachidonic acid (AA), eicosapentaenoic acid (EPA)

and docosahexaenoic acid (DHA) are present in animal sources. Arachidonic

acid is most common in meat while EPA and DHA are mainly present in marine

animal sources such as fish, as well as mammals that primarily feed on fish or

other seafood, such as plankton. Currently the average intake of omega-6 fatty

acids in Sweden is about five times as large as that of omega-3 fatty acids

(Becker, 2002). Traditionally, the anti-inflammatory effect of the omega-3 fatty

acids is attributed to the long-chain EPA, which competes with AA in the AA

cascade for eicosanoid synthesis and produces less pro-inflammatory

eicosanoids (Figure 2) (Calder, 2009).

α-linolenic acidALA 18:3 ω-3

stearidonic acid18:4 ω-3

eicosatetraenoic acid20:4 ω-3

eicosapentaenoic acid EPA 20:5 ω-3

docosapentaenoic acid DPA 22:5 ω-3

docosahexaenoic acidDHA 22:6 ω-3

linoleic acidLA 18:2 ω-6

γ-linolenic acid18:3 ω-6

dihomo γ-linolenic acidDGLA 20:3 ω-6

arachidonic acidAA 20:4 ω-6

docosatetraenoic acid22:4 ω-6

docosapentaenoic acid22:5 ω-6

Δ6 desaturase

elongase

Δ5 desaturase

elongase

Omega-3 family Omega-6 family

Less inflammatoryeicosanoids

elongaseΔ6 desaturaseβ-oxidation

Less inflammatoryeicosanoids

More inflammatoryeicosanoids

Figure 2 The endogenous elongation of shorter dietary polyunsaturated fatty acids and

their resulting eicosanoids.

12

It has been suggested that the body produces a substantial portion of its

requirements of AA from LA (Ratnayake, 2009). This endogenous production

of AA appears to be limited by the activity of delta-5-dehydrolase (D5D) which

may be inhibited by high dietary intake of long-chain omega-3 PUFA's

(Barham, 2000). The human body can produce long-chain omega-3 PUFAs

from ALA by the same pathway as LA. However, this process is limited, since

ALA is found in much smaller quantities in the diet compared with LA. The

human body also prefers to use ALA for energy metabolism by β-oxidation,

which further decreases the capacity to endogenously produce longer omega-3

fatty acids (Cunnane, 1997). To summarise it would therefore be of interest to

study the intake of fats and fatty acid metabolism in AS, especially since

epidemiological studies on native populations with high intake of omega-3 fatty

acids report an increased prevalence of cases with mild disease (Boyer, 1994).

Vitamin D is described to be depleted in AS and PsA when compared with

controls, but its role in the pathogenesis of the respective disease is not fully

understood and results are partially conflicting. Coeliac disease is not more

prevalent in AS, although several studies point to elevated antibodies against

gluten among patients with AS. The importance of dietary fibres has been

discussed, although generally assumed to be beneficial for health; they may also

affect the lumen barrier, as well as having osmotic effects in the bowel.

13

AIMs

General aim

The general aim of the present thesis was to investigate the influence of diet in

AS.

Specific aims

• to investigate the relationship between diet and disease activity among

patients with AS (Paper I and II);

• to investigate whether patients with AS experience gastrointestinal

complaints and if these are related to diet (Paper I);

• to investigate the fatty acid composition in the diet, in gluteal adipose tissue

and in plasma phospholipids, and their correlation with disease activity

among patients with AS (Paper II);

• to investigate whether patients with AS exhibit levels of vitamin D and

transglutaminase outside the normal laboratory limits (Paper II);

• to investigate if diet and some traditional risk markers for CVD among AS

patients differ in comparison to the general population (Paper III); and

• to investigate if a raised intake of long-chained omega-3 fatty acids can

decrease disease activity among patients with AS (Paper IV);

14

METHODS

A: Total identified cohort in Västerbotten, n=1661: Participants in Paper I, n=1222: Participants in Paper II, n=663: Participants in Paper III, n=87

B: The cohort in Norrbotten4: Participants in Paper IV n=24

4

3

2

1

A B

Figure 3 The origin of the cohorts studied for Paper I-IV

Subjects and study design

Paper I: By means of careful examination of medical records covering the period

between May 2002 to May 2007 at the Department of Rheumatology in the

county of Västerbotten northern Sweden, 166 patients aged 18-70 years, with a

validated diagnosis of definite AS according to modified New York criteria (van

der Linden, 1984) could be identified (Figure 3). In this paper, 165 (130 males,

35 females) of these patients were sent a self-administrated questionnaire (one

patient of the original 166 was missed due to an administration error). Of the

cohort, 122 patients chose to participate in the study and returned a filled

questionnaire. Of these 122 questionnaires, two were excluded due to the

amount of missing data and nine were excluded due to contemporary IBD,

yielding 111 participants for analysis. Demographic data of the final group is

presented in Table 1. The patients were treated as their own controls, i.e., by

comparing patients with different outcomes.

Table 1 Demographics for subjects included in Paper I-IV

n (male:female)Age, median (range)

Disease duration, median(IQR)BASDAI, median (IQR)BASFI, median (IQR)

Paper I

111 (84:27)50 (25-70)

28 (9)4.1 (1.3)2.6 (1.8)

Paper II

66 (51:15)49 (26-65)

26 (9)3.8 (1.4)2.3 (1.3)

Paper III

Patients89 (68:21)50 (30-60)

---

Controls355 (272:83)

50 (30-61)---

Paper IV

24 (15:9)49 (33-69)

19 (15)5.1 (2.3)4.0 (2.6)

15

Paper II: From the original cohort studied for Paper 1, patients who were not

pregnant, lactating, prescribed lipid-lowering treatment, had been treated with

dalteparin sodium, warfarin or biological treatment, such as TNF-α inhibitors,

during the past 3 months, were asked to participate and donate blood and

adipose tissue samples for a study of identified biomarkers. Sixty-six patients

agreed to participate and were compared with themselves as controls.

Paper III: The patients (n=166) in Paper I were asked for consent to investigate

whether they had donated data to the Västerbotten Intervention Programme

(VIP), and 148 agreed to participate; of these, 89 (60%) were found in the VIP

database. The VIP is a continuing community intervention programme on

diabetes and CVD, started in Norsjö during 1985, and was gradually extended to

include the total county of Västerbotten in 1991 (Weinehall, 2001). The

database contains health-screening data and blood samples from 40-, 50- and

60-year old citizens (before 1996 30-year olds were also included). The rate of

participation in the programme has been 57%, and the differences in social

characteristics between participants and non-participants has been described as

being small (Weinehall, 1998). The included patients were compared with 355

controls matched for age (±2.5 years), sex and time studied (±2months). The

patients were mainly compared with their matched controls, but correlations

and comparisons were also made within the patient group.

Paper IV: Twenty-four patients with AS participated to a 21-week omega-3 fatty

acid supplementation intervention study. The patients were recruited from the

two hospitals of Gällivare and Kiruna in the iron ore district of far northern

Sweden. Inclusion criteria were: active disease, age 18-70 years and having a

diagnosis of AS made by a rheumatologist. Exclusion criteria were: treatment

with methotrexate, etanercept or infliximab, pregnancy, active Crohn’s disease

or food intolerance to fish. The participants were randomly assigned to either a

low-dose group or a high-dose group receiving 1.95 grams of capsulated long-

chained omega-3 fatty acids per day, or 4.55 grams, respectively. The efficacy

was evaluated using the BASDAI, BASFI and ESR measurements. As safety

precautions, complete blood count (CBC), concentrations of hemoglobin (Hb),

aspartate aminotransferase (AST), alanine aminotransferase (ALT) and

16

gammatransferase (GT) were analysed at the visits to the relevant clinic.

Eighteen patients completed the study period.

Assessments

Table 2 Assessments performed a in Paper I-IV

Assessment of disease

In Paper I, II and IV disease activity and functional capacity were assessed by

the Swedish versions of the self-administered indices BASDAI (Table 2)

(Waldner, 1999) and BASFI (Cronstedt, 1999). In Paper II and IV the

BASDAIBASFIESRhs-CRPSerum triglyceridesSerum cholesterolLipoproteinsCytokinesComplete blood count

Serum MMP-3

Plasma phospholipid composition

Gluteal adipose tissue compositionSerum phosphate S-25-dihydroxyvitamin D (Calcidiol)Serum transglutaminaseSerum IgASerum ALTSerum ASTSerum GTFFQ-84 with backgroundFFQ 64-66Other FFQDrug consumptionGastric complaintsPhysical activity levelWeight & Length, patient reportedWeight & Length, measuredWaist and Hip Circumstance, measuredBlood pressure*not used for nutritional calculations in this thesis

Paper I

xx

x

xxxx

Paper II

xxxxxxxxx

x

x

xxxxx

x

xxxxxxx

Paper IIIPatients

xx

x*x

x

x

x

x

Controls

xx

x*x

x

x

x

x

Paper IV

xxx

x

xxx

x*x

17

assessment of disease activity was complemented with ESR measurements, and

in Paper II it was further complemented by measurement of high sensitive c-

reactive protein (hsCRP), MMP-3 and analysis of cytokines (interleukin-1β

(IL-1ß), IL-1ra, IL-6, IL-17, IFN-γ, MCP and TNF-α).

Assessment of dietary intake

In Paper I and II, dietary habits were assessed by a validated 84-question semi-

quantitative FFQ (Johansson, 2002; Wennberg, 2009). In Paper III, a slightly

shorter version with between 64 and 66 questions were used on subjects

included in the VIP after 1996. The food intake frequency was reported on a

nine level scale ranging from never to four times per day or more. The

questionnaire contained examples of portion sizes on four different plates

regarding vegetables, meat or fish and food staples such as rice, pasta or

potatoes. The frequencies and sizes of portions from the FFQ were used to

calculate the energy distribution, fatty acid distribution and fibre content of the

food. Standard values in the reference databases were used for those foods

where the examples depicted on plates were not valid for estimating portion

size. The calculations were based on reference databases from the Swedish

National Food Administration (Bergström, 1991). If the patients consumed a

supplementary intake of omega-3 fatty acids, this was added to the nutritional

calculations as containing 19 % EPA, 14 % DHA and 23 % SFA in Paper I and II.

In addition to the calculations for nutritional content, the consumption

frequencies were also converted to a weekly or monthly basis and pooled into

groups to describe food patterns.

From the nutritional content the sum of PUFAs were calculated as the sum of

LA, ALA, AA, EPA and DHA. The sum of omega-3 fatty acids was calculated as

the sum of ALA, EPA and DHA, the sum of long-chain omega-3 fatty acids was

calculated as the sum of EPA and DHA. Docosapentaenoic acid (DPA) was not

included in the nutritional calculations since it is not specified in the nutritional

databases due to its presence at very low levels.

In Paper IV a different FFQ, previously used in a thesis on RA and

Mediterranean diet (Hagfors, 2003), was used to assess if any dietary changes

18

occurred among the participants during the intervention. Beyond that, no

further analysis of this FFQ was performed as part of the present thesis.

As a biomarker for the composition of fat intake during the preceding year,

analysis of fatty acid distribution in gluteal adipose tissue was performed. For

analysis of fat intake composition during the preceding weeks, plasma samples

were taken and analysed for the fatty acid content of phospholipids. The

distribution of fatty acids from these analyses was calculated as the mass of total

fatty acids, i.e., SFA's, PUFAs, long-chained PUFAs, omega-3 and omega-6 fatty

acids. Polyunsaturated fatty acids were calculated as the sum of LA, ALA, AA,

EPA, DPA and DHA fatty acids. Long-chained PUFAs were calculated as the

sum of AA, EPA, DPA, DHA fatty acids. Omega-3 fatty acids were calculated as

the sum of ALA, EPA, DPA, DHA fatty acids and omega-6 fatty acids as the sum

of LA and AA.

From the laboratory analysis of fat distribution, desaturase activity was

estimated by calculating the ratio between measurable products and their

measurable precursors in plasma phospholipids. For Stearoyl-CoA desaturase

(SCD) this resolved to a ratio as 16:1n7/16:0 and for D5D

20:5n3+20:4n6/20:3n6, since levels of 20:4n3 are low due to the preference of

desaturation of omega-3 fatty acids for D5D. Delta-6-desaturase (D6D) was not

calculated due to low levels of its products 18:3n6 and 18:4n3.

Background factors and other specified assessments

Level of education, social situation, smoking history, height and body weight

were assessed by questions from the population studies in northern Sweden in

Papers I-III (Stegmayr, 2003). The patients were classified as smokers when

reporting an average consumption of one or more cigarettes per day. Use of any

pharmacological therapy for AS disease was investigated by an open question on

specification of type and frequency used in Paper I, II and IV. In Paper III

medications were assessed in questions specifically asking about drugs used for

high blood pressure, heart disease or angina pectoris, sleeping problems, gastric

ulcers and/or hyperlipidaemia during the preceding 14-day period. The

19

presence of coeliac disease was screened for in Paper II by conventional hospital

routines with analysis of transglutaminase and serum IgA in blood samples.

A physical activity level (PAL) was estimated in Paper I, II and III. The

assessment was based on two questions regarding activity levels at work and

during leisure time which result in a calculated value for PAL (Johansson,

2008). This calculated PAL value was validated against doubly labelled water in

a small pilot study, showing a mean difference between the methods of only

0.004 (0.172).

In Paper I the presence of gastrointestinal problems was investigated by

questions such as: ‘Do you have problems with hard stools or constipation?’, ‘Do

you have problems with loose stools?’, and ‘Do you have problems with stomach

ache or other gastrointestinal pain?’. The answers to these questions had five

options ranging from ‘no/never’ to ‘several times a day’. In open questions, the

patients were also asked if any particular foodstuff(s) aggravated either

gastrointestinal difficulties or the symptoms of AS disease.

In Paper II, anthropometric assessment was performed by weighing the patients

with light clothing and without shoes using a digital calibrated scale (Seca Delta

model 707). Height was measured using a wall mounted tape measure, and

waist and hip circumference by using a regular tape measure (Kallings, 2002).

In Paper II, blood pressure was measured using a digital cuff (Bosch boso

medicus CA01) in a sitting position following 10 minutes rest. In Paper III,

blood pressure was measured in sitting position by ordinary primary care

routines. High blood pressure was defined by either systolic pressure greater

than 140 mm/Hg, diastolic pressure greater than 90 mm/Hg or by current

medication with anti-hypertensive pharmaceuticals.

In the intervention study of Paper IV, the patients were asked to estimate the

effect of the omega-3 supplementation on their disease symptoms at each revisit

by marking any of seven alternatives ranging from ‘much worse’ to ‘much

20

better’. They were also given the opportunity to write down in free text any

observation made during the trial.

Laboratory performance and routines

Routine hospital protocols were used for the analysis of CBC, ESR, ALT, AST,

serum calcidiol, serum phosphate, serum transglutaminase, serum IgA, serum

triglycerides (TG), high-density lipoprotein (HDL), low-density lipoprotein

(LDL) and total cholesterol. Frozen blood samples were stored below -70° C for

up to seven months and were later analysed for hsCRP, MMP-3, IL-1β, IL-1ra,

IL-6, IL-17, IFN-γ, monocyte chemotactic protein-1 (MCP-1) and TNF-α using

ELISA arrays. These analyses were performed according to the respective

manufacturer's descriptions.

For analysis of fat distribution in adipose tissue, subcutaneous tissue samples

were taken from the upper left gluteal quadrant of the patients with a

Vacutainer® tube and needle as described by Beynen and Katan (Beynen,

1985). The tissue samples were collected and stored in cryo tubes below -70º C

before further analysis. Blood was sampled into Vacutainer® tubes coated with

ethylenediaminetetraacetic acid (EDTA) as anti-coagulant and was centrifuged

at 2500-3000xg for 10 minutes after resting for 30 minutes before freezing the

resultant plasma. Before analyses, both plasma and adipose tissue were stored

up to 7 months at below -70º C. The analyses of fat composition were

performed at the Department of Public Health and Caring Sciences, Unit for

Clinical Nutrition Research, University of Uppsala, Sweden as previously

described (Boberg, 1985).

Ethical considerations

All studies were approved by the Regional Ethical Review Board in Umeå,

Sweden (Dnr 03-207, Dnr 07-082M, Dnr 07-173, Dnr 2010-89-32) and were

performed in concert with the declaration of Helsinki.

Statistical methods

A missing value in either the BASDAI and BASFI was replaced by a calculated

mean value. Missing values in the FFQ were treated as follows in Paper I and II:

21

if less than 10% of the questions were unanswered, those questions were

assigned a value of zero intake, whilst if the questionnaire contained more than

10% unanswered questions the responder was excluded from the study. In

Paper III, subjects with a food intake level (FIL) below the 5th percentile and

above the 97.5 percentile were classified as under- and over-reporters with

inadequate data and were removed from further calculations on diet.

Furthermore, in Paper III, subjects completing the longer FFQ used before 1996

(n=6) were excluded from the calculations on diet to simplify statistical analysis.

To increase robustness against remaining over- and underreporting, nutritional

content was calculated as related to reported energy intake, i.e., as energy

percent or grams per kilojoule (kJ). The results were considered statistically

significant at a two-tailed p-value of ≤0.05. Statistical calculations were

performed with PASW 18.0 for Macintosh for Paper I-II, SPSS 19.01 for Paper

III and with Statview 5.0 for Macintosh in Paper IV.

22

RESULTS AND DISCUSSION

Paper I

Diet, disease activity and gastrointestinal symptoms in patients with

ankylosing spondylitis

Dietary intake is summarised in Table 3. Women had significantly lower intake

of fat and significantly higher intakes of fruits, vegetables, fibre and fish.

Despite a healthier diet, the included women had numerically higher BASDAI

compared with the men.

Table 3 Dietary intake of 111 patients with ankylosing spondylitis. Data are presented

as median (IQR) of dietary intake, calculated as percentage of energy intake (E%) as well

as consumption frequencies.

*p-value<0.05, **p-value<0.01, ***p-value<0.001. P-values denote statistically significant

differences between men and women.

For the group as a whole, no effect on disease activity due to diet was observed.

This was shown both by the correlation estimates between dietary intake and

the BASDAI scale, and by the fact that only seven patients out of the 111

experienced a worsening in their AS disease to be associated with certain foods.

This is lower than previously reported when one-third of the AS patients

reported aggravation of their disease to be related to diet (Haugen, 1991).

However, with the current introduction of high efficacy biological agents, there

Fat, E %Protein, E %Carbohydrate, E %

Milk and soured milk, servings/monthMeat and meat products, servings/monthFish, servings/month

Vegetables, servings/month

Fruit, servings/month

Fiber, grams /MJ

All patients (n=111)

36 (32-40)15 (13-16)45 (41-49)

43 (27-78)23 (19-28)

5.8 (4.3-7.6)

44 (28-61)

33 (18-54)

2.4 (1.9-2.9)

Men (n=84)

38 (34-41)15 (14-16)45 (42-48)

46 (30-80)23 (20-29)

5.8 (4.3-7.6)

38 (23-58)

30 (18-46)

2.3 (1.8-2.9)

Women (n=27)

31 (27-35)***14 (13-16)45 (40-54)

35 (22-68)22 (18-27)

7.6 (4.9-11.0)*

57 (45-81)***

52 (31-95)**

2.8 (2.3-3.1)**

23

is far less focus on diet, which may be a major reason for the lower reporting in

the present study. Nevertheless, among women, a negative correlation was seen

between BASDAI and intake of total fat (rs =- 0.43, p <0.05) as well as between

BASDAI and intake of saturated fat (rs =- 0.50, p <0.01). This result is difficult

to interpret, but in combination with the finding that women had significantly

lower fat intake than men, it could indicate that a low fat intake may have a

negative effect on disease activity in AS.

The effect of diet on perceived gastrointestinal discomfort was more explicit.

More than a quarter of the patients associated consumption of certain foods to

the presence of gastrointestinal discomfort. The reported foods, i.e., dairy

products (n=7), vegetables/fruit (n=7), fatty foods (n=6) and food rich in flour

(n=5) resembles the responses from patients with IBD as to the food stuff

affecting gastric symptoms (Ballegaard, 1997; Jowett, 2004). There were

significantly higher estimates on the BASDAI (p<0.01) and BASFI (p<0.05)

scales among patients with gastrointestinal pain. There was also a higher

consumption of vegetables (p=0.01), as well as lower intake of milk (p<0.05)

among patients with gastric pain. The higher estimates on disease scales could

implicate that more severe disease is linked to increased gastrointestinal

problems. Since vegetables are often regarded as healthy, patients with

gastrointestinal symptoms may try to increase consumption in belief that it

would be beneficial. Conversely, milk intolerance is well known within the

general population, hence the patients may exclude or lower their consumption

of milk in attempts to relieve gastrointestinal symptoms. Since AS patients are

reported to have changes in the bowel mucosa that resemble those seen in IBD

(De Vos, 1989: Mielants, 1991), it is reasonable that diet induced gastric

symptoms would be quite similar to those seen in IBD. In both IBD and the

irritable bowel syndrome (IBS) it has been suggested that an impaired intestinal

breakdown of carbohydrates, such as flour products, vegetables and lactose,

would cause gastric symptoms when they are unabsorbed when reaching the

ileocolonic junction and become substrate for rapid bacterial growth (Vonk,

2003; Gibson, 2005).

24

A high prevalence of gastrointestinal discomfort was seen among the patients.

This was seen irrespective of whether the patients had consumed a NSAID or

not (Figure 4). These findings are contradictory to the traditional view of

intestinal inflammation in patients with AS being clinically silent (Rudwaleit,

2006). Similar viewpoints have recently been raised by Stebbings et al. who

showed that patients with active SpA have bowel symptoms similar in severity

to that experienced by patients with Crohn’s disease (Stebbings, 2009).

0%

15%

30%

45%

60%

Gastric pain Loose stools Constipation

NSAID No NSAID

Figure 4 Gastrointestinal complaints reported by 111 patients with ankylosing

spondylitis, stratified for consumption of NSAID

Although the consumption of NSAIDs is often reported to cause an upset

stomach, patients consuming traditional NSAIDs did not report more gastric

symptoms than non-consumers. One explanation could be that patients with

gastrointestinal problems tend to avoid NSAIDs due to their gastrointestinal

side-effects; another explanation is that NSAIDs also may relieve

gastrointestinal pain.

Conclusions and implications: Diet within the dispersion normally seen in

western populations does not affect disease activity in AS, but may influence

associated problems, such as gastrointestinal pain. Gastric pain is not

uncommon in AS especially among patients with greater disease activity.

25

Paper II

Plasma phospholipids are related to disease activity in ankylosing

spondylitis

Consistent with Study I, no apparent correlation between dietary intake and

disease activity assessed by BASDAI was identified. There were, however,

negative correlations between ESR and intake of PUFAs (rs=-0.25, p<0.05) as

well as long-chain omega-3 PUFAs (rs=-0.27, p<0.05). Since this was not

reflected with hsCRP levels, or other disease activity parameters, it may be

questioned whether this is an effect of other factors such as blood fluidity rather

than an effect on inflammation (Cartwright, 1985).

Table 4 Correlation between plasma phospholipid content of AA, DGLA, EPA, D5D and

disease activity according to BASDAI and the six separate BASDAI questions among 66

patients with ankylosing spondylitis.

*p<0.05 **p<0.01

The main finding in study II was that plasma phospholipid content of AA

correlated significantly with the BASDAI score (rs=0.39, p<0.01; table 4). An

increased calculated D5D activity was also found, suggesting that there may be

an enhanced biosynthesis of AA with a consequent increased incorporation into

phospholipids, which may be related to a higher disease activity in patients with

AS. This is particularly interesting bearing in mind that none of the other

biomarkers analysed, i.e., ESR, hsCRP, MMP-3, IL-1β, IL-1ra, IL-6, IL-17, IFN-

γ, MCP and TNF-α, correlated with BASDAI. Plasma AA may, therefore, be a

potential biomarker for disease activity, reflecting a process involved in the

inflammation associated with AS. This finding is of considerable interest since a

AA

DGLA

EPA

D5D

BASDAI

0.39**

-0.13

0.07

0.37**

Fatigue

0.26*

0.02

0.09

0.22

Back pain

0.40**

-0.12

0.05

0.35**

Pain/swelling

peripheral joints

0.13

-0.09

0.02

0.16

Enthesitis

0.29*

-0.16

0.14

0.33**

Severity morning stiffness

0.41**

-0.18

0.08

0.40**

Duration morning stiffness

0.30*

-0.13

0.01

0.31*

26

hallmark of AS is the good symptom relief achieved with NSAIDs, a medication

that inhibits the synthesis of pro-inflammatory eicosanoids from AA (Yu,

2000). As yet little is known about which mechanisms control either

endogenous production of long-chain PUFAs or their incorporation in

phospholipids. Since there is an obvious increased utilisation of AA from

phospholipid membranes in inflammatory conditions, such as AS, there must be

mechanisms present to swiftly replace the lost AA. Otherwise, membrane

phospholipids would soon be lacking a sufficient amount of AA and the

inflammatory cascade would cease.

Table 5 Correlations between dietary intake, calculated as percentage of energy intake

(E%) as well as consumption frequencies, and atherogenic lipids among 66 patients with

ankylosing spondylitis.

*p<0.05 **p<0.01. Spearman rank correlation test used for correlations

Regarding blood lipids involved in atherogenic processes, correlations between

HDL cholesterol and the intake frequency of both fish, and vegetables and fruit

were seen (Table 5). In addition, a high intake of fish was also associated with

lower plasma levels of TG. It has previously been well described how lifestyle

factors, such as diet, can influence atherogenic blood lipids in the general

population (Lairon, 2007; Sanders, 2009). Since an increased cardiovascular

morbidity has been described in AS (Han, 2006; Bremander, 2011) these

findings may have implications for cardiovascular preventive strategies in AS.

Dietary intake

Fat, E%Protein, E%Carbohydrate, E%Saturated fatty acids, E%Linoleic acid, E%Alpha-linolenic acid, E%Polyunsaturated fatty acids, E%Long-chained omega-3 fatty acids, E%

Milk and soured milk, servings/monthMeat and meat products, servings/monthFish, servings/monthVegetables, servings/monthFruit, servings/month

median (IQR)36.0 (32.3-39.0)14.7 (13.6-15.6)46.0 (42.1-49.6)14.4 (12.6-16.4)

4.3 (3.4-5.3)0.85 (0.68-1.03)

5.4 (4.3-6.5)0.13 (0.09-0.18)

76.9 (37.4-99.6)22.2 (17.1-30.4)

4.5 (2.2-8.4)38.9 (23.6-57.2)38.5 (20.7-67.8)

Correlations to atherogenic lipid factors

TG -0.36**

-0.01 0.38** -0.38**

-0.05-0.10 -0.08

-0.25*

-0.44**-0.21

-0.26*0.07-0.01

Cholesterol 0.15

-0.00 -0.30*

-0.05 0.22 0.08 0.21

0.25*

−0.21 0.04 0.20 0.14 0.09

HDL 0.16-0.14

-0.33** 0.16-0.06-0.09-0.03

0.26*

0.220.16

0.40** 0.35** 0.31*

27

Conclusions and implications: The correlation between plasma

phospholipid content of AA and disease activity may reflect a process involved

in the inflammation associated with AS. Arachidonic acid is, therefore,

identified as a potential and feasible biomarker for inflammatory activity. Diet

is correlated to atherogenic blood plasma lipid levels and may affect

cardiovascular risks.

Paper III

Diet and cardiovascular risk factors among patients with ankylosing

spondylitis in comparison with the general population

In this paper the main finding was the divergent lipid profile among the patients compared with the general population (Table 6). In accordance with what was initially described by Masi et al (Masi, 1999; Masi, 2000), we found highly significant differences on cholesterol and TG between patients and matched controls were observed despite the relatively low disease activity (described in Papers I and II).

Table 6 Clinical data for 89 patients with ankylosing spondylitis and 355 matched

controls.

Exercise, n/week. median (IQR)

PAL, median (IQR)

Smokers, n (%)Triglycerides (mmol/L) median (IQR)Cholesterol (mmol/L) mean (SD)

High blood pressure mean (SD)

Pharmacotherapy

Anti-hypertensive

Cardioprotective

Sedatives

Ulceroprotective

Lipid-lowering

Cases (n=89)

2 (1-4)

1.6 (1.5-1.8)

14 (16.5%)

1.08 (0.80-1.44)

5.11 (1.05)

26 (31.3 %)

18 (21.4 %)

1 (1.2 %)

3 (3.6 %)

14 (16.7 %)

4 (4.8 %)

Controls (n=355)

2 (1-3)

1.6 (1.5-1.8)

44 (12.1%)

1.21 (0.84-1.77)

5.49 (1.21)

95 (29.0 %)

44 ( 13.4 %)

13 (3.9 %)

15 (4.6 %)

19 (5.8 %)

14 (4.3 %)

p-value

p=0.26

p=0.47

p=0.41

p<0.01

p<0.01

p=0.74

p=0.09

p=0.21

p=0.61

p<0.01

p=0.86

28

Despite physical activity being emphasised as a cornerstone in the treatment of

AS (Zochling, 2006), half of the patients exercised on fewer than two occasions

a week (data not shown). Furthermore, no differences between patients and

controls were shown either using PAL or number of exercise sessions per week

(Table 6). This puts stress on the need for research on compliance to the

exercise regimes advocated for patients with AS.

Table 7 Correlations between background data and dietary intake, calculated as

percentage of energy intake and atherogenic lipids among 77 patients with ankylosing

spondylitis and 307 controls.

*p<0.05 **p<0.01

Analysing diet and atherogenic lipids revealed that the level of triglycerides

among the patients was inversely correlated to the intake of total fat (rs = -0.25,

p<0.05; Table 7), and monounsaturated fats (rs = -0.29, p<0.05) whilst they

showed a positive correlation to the intake of carbohydrates (rs = 0.26, p<0.05).

Background data and dietary intake

BMIPALExercise

Fat

Protein

Carbohydrate

Saturated fatty acids

Monounsaturated fatty acidsPolyunsaturated fatty acidsLinoleic acid

Alpha-linolenic acidLong-chained omega-3 fatty acids

Cases (n=77)

Correlations to atherogenic lipid factors

TG (mmol/L)

0.30**0.120.10

-0.25*

-0.14

0.26*

-0.21

-0.29*-0.14-0.09

-0.17

-0.02

Cholesterol (mmol/L)

-0.190.010.02

-0.20

-0.21

0.26*

-0.10

-0.28*-0.15-0.13

-0.16

-0.17

Controls (n=307)

Correlations to atherogenic lipid factors

TG (mmol/L)

0.38**-0.09-0.08

0.01

0.01

-0.02

0.03

0.03-0.06-0.06

-0.09

0.05

Cholesterol (mmol/L)

0.08-0.09-0.08

-0.10

-0.07

0.10

-0.07

-0.14*-0.10-0.08

-0.06

0.05

29

Similar findings have been described in studies on larger, general populations

(Hu, 2002), however the correlations were more pronounced among the

patients than in the controls studied. This suggests a stronger influence of diet

on blood lipids in patients with AS than within the general population.

The calculated energy intake was significantly higher among patients than

controls (mean 1904 kcal, compared with 1819 kcal; p<0.05). This difference

remained after adjusting for PAL, weight and education level in multiple

regression analyses (data not shown). As increased resting energy expenditure

(REE) has been described in inflammatory conditions (Roubenoff, 1994; Staal-

van den Brekel, 1995) this may implicate an increased REE also in AS. Together

with the present findings of depressed levels of blood lipids, this may suggest

differences in metabolism among AS patients compared with the general

population.

Conclusions and implications: Patients with AS exhibit significantly lower

atherogenic blood lipids compared to controls. Patients with AS do not differ

significantly from the general population regarding lifestyle factors such as

physical activity, smoking and diet, but there appears to be a more pronounced

effect of diet on blood lipids among the patients.

Paper IV

Supplementation of Omega-3 fatty acids in patients with ankylosing

spondylitis

In this study it was found that any form of treatment with omega-3 requires

both a long duration and high doses to be efficient. The lower daily dose of 1.95

grams did not generate any demonstrable effect on disease activity based on

BASDAI (Table 8). In the high dose group however, a significant reduction was

observed on the BASDAI score, but only after 21 weeks. Improvement, defined

as a reduction in BASDAI by at least one-third was apparent among six of the

nine patients in the high dose group, compared with two of nine patients in the

low dose group. When comparing the low- and high-dose groups, there was a

30

greater numerical improvement in BASDAI in the high-dose group, although

this did not reach a statistically significant level (p = 0.15).

Table 8 Disease activity, functional index and safety assessments during a 21-week

intervention study with omega-3 supplementation in ankylosing spondylitis.

*4.55 grams of long-chain omega-3 fatty acids per day **1.95 grams of long-chain omega-3 fatty acids per day. P-

values denote statistically significant differences between baseline and end of study at 21 weeks. ns: not significant

The safety of the supplementation was evaluated in terms of ALT, AST, GT, Hb

and CBC. There was a significant reduction of ALT in the whole group (p<0.01;

data not shown) as well as for the low-dose sub-group (p<0.05). This may

indicate that supplementation with omega-3 fatty acids causes reduced burden

on the liver, since a large part of the body's endogenous production of long-

chain PUFAs, as well as its general fatty acid metabolism, occurs there.

Side-effects that were reported during the supplementation were abdominal

discomfort (n=6), which led to two participants suspending their participation.

In total, 6 patients discontinued the study. Other than gastric complaints (n=2),

the reasons were bad compliance, lack of motivation, anaemia and pneumonia,

respectively.

Omega-3 supplementation may be an alternative to conventional NSAID's

treatment in AS. Since the dose in this study is higher than that usually used in,

for example, CVD (Psota, 2006), more extensive studies on the efficacy, side-

effects and health economics are needed. Converted to a dietary intake, 4.55

grams/day, equals almost two servings of oily fish per day, e.g., 100 grams of

fried or cooked salmon equals 2.6 or 2.7 grams of long-chained omega-3 fatty

High-dose group* (n=9)

BASDAI median (IQR )

BASFI median (IQR )

Low-dose group** (n=9)

BASDAI median (IQR )

BASFI median (IQR )

Baseline

4.32 (1.32)

3.14 (2.95)

3.01 (2.39)

3.83 (1.91)

7 weeks

2.62 (1.92)

2.42 (2.55)

3.00 (1.58)

4.78 (2.67)

14 weeks

3.37 (1.12)

2.94 (2.80)

2.72 (2.15)

4.40 (5.82)

21 weeks

2.92 (1.42)

2.55 (1.76)

2.44 (2.99)

4.15 (3.56)

p-value

p<0.05

ns

ns

ns

31

acids. Although it has been suggested that fatty acids supplied as fish yield a

slightly higher bioavailability than the corresponding amount supplied as

capsules (Harris, 2007), in practical use and implemented to daily routine, it

will probably not be easy to reach therapeutic levels of long-chain omega-3

without the use of supplements.

Conclusions and implications: High-dose omega-3 supplementation

lowers disease activity in AS, as measured by BASDAI, and may, therefore, be an

alternative to conventional treatment such as NSAID's. Due to the high doses

required, further studies should be performed in order to evaluate long-term

efficacy and side-effects.

General discussion, limitations and future perspective

From these studies it is concluded that diet, within the normal range of

consumption in the western countries has little, or no effect, on disease activity

in patients with AS. However, it is important to emphasise that this does not

preclude the possibility that a radical diet, such as an Inuit diet, can affect the

course of the disease as implicated by the results from the high-dose omega-3

intervention in presented in Paper IV. Studies on Alaskan Inuit by Boyer et al.

are interesting in this context as these individuals, despite a high incidence of

HLA-B27, showed an unexpectedly low frequency of spondyloarthropathies,

particularly AS, as well as the remarkably mild nature of the cases found (Boyer,

1994). Although Boyer and colleagues did not investigate diet, other studies on

similar and related Inuit groups report a diet containing a low percentage of

carbohydrate, with virtually non-existent amounts of fruit, vegetables and dairy

products, and a high proportion of protein and fat, especially MUFA and omega

-3 fatty acids (Sinclair, 1953; Bang, 1980).

Consistent with the results reported in Paper IV, radical changes in fat intake, or

high-dose supplementation, have been argued to achieve effect in chronic

inflammatory diseases (Cleland, 2006; Galli, 2009). There are side-effects

associated with the common pharmacological agents used in the treatment of

AS; with NSAID's these are gastrointestinal and cardiovascular side-effects

(Peterson, 2010), whereas TNF-α inhibitors give rise to a risk of infection or

32

hypersensitivity reactions (Rosenblum, 2011). High-dose omega-3

supplementation may, therefore, fill a void as a therapeutic option for patients

with AS intolerant to NSAID's and/or not eligible for treatment with TNF-α

inhibitors. It has also been shown that supplementation with long-chain

omega-3 PUFAs has a positive effect on heart conduction defects (Richardson,

2011). Therefore, supplementation is also interesting with regard to the

conduction disorders that may be present in AS (Bergfeldt, 1982; Momeni,

2011).

The finding in Paper II of a correlation between disease activity and the amount

of AA in plasma phospholipids emphasise the importance of long-chain PUFAs

metabolism in AS. An intake of omega-3 within the normal range of a western

population did neither affect the disease activity nor the deposition of AA in

plasma phospholipids. Since man may form the necessary AA from LA

(Ratnayake, 2009) diet will have to outnumber LA with ALA to some degree to

reach an anti-inflammatory effect. Alternatively, if the diet contains a very high

amount of EPA, it can compete with AA in the production of eicosanoids

(Calder, 2006) as well as inhibit the endogenous production of AA (Barham,

2000). In recent years, there have been suggestions that polymorphisms of the

FADS-1 and -2 genes that encode production of long-chain PUFAs may raise

levels of AA, and thereby contribute to both CVD and inflammatory diseases

(Martinelli, 2008; Merino, 2010). Therefore, further studies on genes encoding

these desaturases should make an interesting contribution to our

understanding.

If endogenous production of long-chain PUFAs has a role in the etiology of AS,

it is noteworthy that the activity of D5D may be inhibited. It has been shown

that sesamin, a lignin non-fatty acid part of sesame oil, has an anti-

inflammatory effect, most probably through the ability to inhibit the activity of

D5D on omega-6 fatty acids, i.e., the conversion from DGLA to AA (Shimizu,

1991; Chavali, 1998; Utsunomiya, 2000). There have also been attempts to

inhibit D5D activity through pharmaceutical intervention (SC-26196), however

this has not, to the best of my knowledge, been tested on humans (Obukowicz,

1998).

33

With reference to the increased prevalence of CVD in patients with AS, it is

important to note the significantly lower levels of triglycerides and cholesterol

present among the patients possibly leading to an underestimation of the risk

for CVD. As of today, little is known as to the causes of the increased prevalence

of CVD in AS and further studies are needed into this topic. In the general

population, several intervention studies show that diet has a substantial impact

both on cardiovascular risk factors, such as an atherogenic lipid profile, and on

cardiovascular events (Hu, 2002). Among the patients studied pronounced

correlations between diet and lipid profile were found, indicating that a dietary

intervention may have a positive effect on the lipid profile. However, since

these levels are depressed with higher disease activity there may be some

concerns as to whether further suppression of triglyceride and/or cholesterol

levels would be a desirable goal (van Halm, 2006).

The main aim of this thesis was to investigate the relationship between diet and

disease activity in AS. A healthy control group was not considered necessary to

achieve this aim in either Paper I or II. However, it would have been of value to

compare gastric symptoms, and results regarding AA and D5D in subjects

without AS. In the cross-sectional studies, a relatively large and comprehensive

cohort of patients with a verified diagnosis of AS was included. In these studies

semi-quantitative FFQs were used to investigate dietary intake. Although, a

DHI may have yielded more precise data on fat intake, it was not a viable

alternative for practical and economical reasons due to the relatively large size

of our cohorts. However, the FFQ used has been validated with good results

regarding omega-3 long-chain PUFAs (Wennberg, 2009). When examining

data collected in the FFQ for Paper I and II, there were few answers missing.

Therefore only participants with more than 10 % missing answers (two

participants in Paper I, and none in Paper II) were excluded. To add robustness

against under- and over-reporting, diet was calculated as proportions of energy

intake. In Paper III, the same approach was used, but instead of excluding

patients on basis of missing answers, the top 2.5 % and bottom 5 % of the

material based on FIL-value was omitted. Lastly, in Paper IV a low-dose group

was used instead of a traditional control group, partly because no previous study

has described appropriate dosages in this particular disease, and partly because

34

of the lack of a suitable placebo oil at the time of the study. One cannot ignore

the possibility that patients in the low-dose group were, in fact, affected by the

supplementation, and thereby diluted the differences in response between the

groups.

To summarise, the topic of diet and fatty acid metabolism in patients with AS

deserves closer attention. Although diet did not have a direct effect on disease

activity, its effect on blood lipids may be relevant for the prevention of future

CVD among AS patients. Therefore, studies aimed at investigating the role of

lifestyle factors, such as diet, on cardiovascular risk profiles should be

addressed. The findings of increased plasma phospholipid content of AA with

increased disease activity calls for further investigations into fatty acid

metabolism in AS. Such studies should aim at the mechanisms behind the

incorporation of fatty acids into phospholipids and the endogenous production

of long-chain PUFAs, such as the genetics of FADS-1 and -2.

35

CONCLUSIONS

◊ In a group of Swedish AS patients, no correlation between diet and disease

activity could be detected. This does not preclude the possibility that a

radical shift in diet may influence disease activity.

◊ There were correlations between diet and gastrointestinal pain.

Gastrointestinal problems were also found to be prevalent in AS,

independent of NSAID usage.

◊ There was a positive correlation between levels of AA in plasma

phospholipids and disease activity assessed by BASDAI in patients with AS.

This may reflect a process involved in the inflammation associated with AS.

◊ Diet was correlated with atherogenic blood lipid levels and, therefore, may

affect the risk of cardiovascular disease. There may be a more direct effect

from the diet on lipid levels among AS patients than in the general

population.

◊ Patients with AS did not differ from the general population regarding

lifestyle factors such as diet, physical activity and smoking. However, they

exhibited significantly lower levels of cholesterol and triglycerides compared

with the general population, which may lead to an underestimation of risk

for cardiovascular disease in AS.

◊ Omega-3 fatty acids in adequate doses may have the capacity to decrease

the disease activity of AS. High-dose omega-3 supplementation may,

therefore, fill a void as a therapeutic option in AS for patients not

considered suitable for treatment with NSAID's or TNF-α inhibitors.

36

ACKNOWLEDGEMENTS

I wish to express my gratitude to all of the people who have inspired and

supported me during the preparation of this thesis:

• Associate professor Solveig Wållberg Jonsson, my supervisor, for your

kindness, your enthusiasm and your never-ending patience with me

• Professor Gunnar Johansson, my co-supervisor, for your knowledge of

nutrition, for introducing me to the field of nutritional research and for

believing in me

• Professor Solbritt Rantapää Dahlqvist, for allowing me to perform this

research and to have created such an innovative research environment at the

department

• Agneta Uddhammar and Gerd-Marie Alenius, former and present head of the

Rheumatology Clinic, for allowing me to undertake the research reported in

Paper I-III and for providing me with room and shelter

• Linda Hagfors, co-author, for giving me inspiration, skilful supervision and

guidance at the beginning of my nutritional research

• Professor Tommy Cederholm, co-author, for valuable comments and

discussions in the field of fatty acids

• Professor Ingegerd Johansson, co-author, for help with nutritional

calculations and for skilful comments

• Kjell Stålnacke, co-author, for taking the time and giving me the opportunity

to perform my first study in the far north of Sweden

• Heidi Kokkonen, co-author, for valuable comments and for helpful assistance

with the laboratory analyses

• Catharina Eriksson, Lisbeth Söderlund and Solveig Linghult for helpful

assistance and guidance with the laboratory analyses

• Lotta Ljung, my PhD student colleague, for our discussions in science,

medicine and philosophy, and for always cheering me up

37

• Åsa Ågren, Anna-Sara Molin and all the other people at Northern Sweden

Medical Research Bank for help in gaining access to the data of Paper III

• Sonja Odeblom for skilful handling of all mailings, reminders, and consent

forms to the patients

• AnnCathrin Kallin for great support, and for making the sampling and

physical examinations of the patients in Paper II such a joy

• Brian Ellis for skilful proof-reading and professional comments

• My physiotherapy colleagues at the Department of Rheumatology for your

interest and your encouragement

• All the other people at the Department of Rheumatology for always making

me feel welcome and supported at the clinic

• Anders, Peter, Anna, Hans and Kent, as well as my other friends, for

reminding me that there is actually a life besides writing a thesis

• Steve Jobs, whose creativity and ingenuity has influenced areas far beyond

technology - including this thesis

• Lastly, but not the least, my greatest gratitude to all the participating

patients, whose time and effort made this research possible

This work was supported by grants from the northern county councils Visare

Norr, the Borgerskapet fund in Umeå, the Department of Research Norrbotten

county council, the Swedish Rheumatism Association and the JC Kempe

Memorial Scholarship Fund, the Anna Cederbergs foundation for medical

Research and from the European society for clinical nutrition and metabolism.

38

REFERENCES

Adelizzi, R. A., Pecora, A. A., & Chiesa, J. C. (1982). Celiac disease: case report

with an associated arthropathy. Am J Gastroenterol, 77(7), 481-485.

Appelboom, T., & Durez, P. (1994). Effect of milk product deprivation on

spondyloarthropathy. Ann Rheum Dis, 53(7), 481-482.

Appelrouth, D., & Gottlieb, N. L. (1975). Pulmonary manifestations of

ankylosing spondylitis. J Rheumatol, 2(4), 446-453.

Arab, L., & Akbar, J. (2002). Biomarkers and the measurement of fatty acids.

Public Health Nutr, 5(6A), 865-871.

Ballegaard, M., Bjergstrom, A., Brondum, S., Hylander, E., Jensen, L., &

Ladefoged, K. (1997). Self-reported food intolerance in chronic

inflammatory bowel disease. Scand J Gastroenterol, 32(6), 569-571.

Bang, H. O., Dyerberg, J., & Sinclair, H. M. (1980). The composition of the

Eskimo food in north western Greenland. Am J Clin Nutr, 33(12),

2657-2661.

Barham, J. B., Edens, M. B., Fonteh, A. N., Johnson, M. M., Easter, L., &

Chilton, F. H. (2000). Addition of eicosapentaenoic acid to gamma-

linolenic acid-supplemented diets prevents serum arachidonic acid

accumulation in humans. J Nutr, 130(8), 1925-1931.

Barrett-Connor, E. (1991). Nutrition epidemiology: how do we know what they

ate? Am J Clin Nutr, 54(1 Suppl), 182S-187S.

Becker, W., & Pearson, M. (2002). Riksmaten 1997-98: kostvanor och

näringsintag i Sverige: metod-och resultatanalys. Livsmedelsverket.

Bergfeldt, L. (1997). HLA-B27-associated cardiac disease. Ann Intern Med,

127(8 Pt 1), 621-629.

Bergfeldt, L., Edhag, O., & Vallin, H. (1982). Cardiac conduction disturbances,

an underestimated manifestation in ankylosing spondylitis. A 25-year

follow-up study of 68 patients. Acta Med Scand, 212(4), 217-223.

Bergfeldt, L., Insulander, P., Lindblom, D., Moller, E., & Edhag, O. (1988). HLA-

B27: an important genetic risk factor for lone aortic regurgitation and

severe conduction system abnormalities. Am J Med, 85(1), 12-18.

39

Bergström, L., Kylberg, E., Hagman, U., Erikson, H., & Bruce, Å. (1991). The

food composition database KOST: the National Food Administration’s

Information System for nutritive values of food. Vår Föda, 43, 439-447.

Beynen, A. C., & Katan, M. B. (1985). Rapid sampling and long-term storage of

subcutaneous adipose-tissue biopsies for determination of fatty acid

composition. Am J Clin Nutr, 42(2), 317-322.

Bjelle, A., Cedergren, B., & Rantapää Dahlqvist, S. (1982). HLA B 27 in the

Population of Northern Sweden. Scandinavian Journal of Rheumatology,

11(1), 23-26.

Bjorkkjaer, T., Brun, J. G., Valen, M., Arslan, G., Lind, R., Brunborg, L. A. et al.

(2006). Short-term duodenal seal oil administration normalised n-6 to

n-3 fatty acid ratio in rectal mucosa and ameliorated bodily pain in

patients with inflammatory bowel disease. Lipids Health Dis, 5, 6.

Bjorkkjaer, T., Brunborg, L. A., Arslan, G., Lind, R. A., Brun, J. G., Valen, M. et

al. (2004). Reduced joint pain after short-term duodenal administration

of seal oil in patients with inflammatory bowel disease: comparison with

soy oil. Scand J Gastroenterol, 39(11), 1088-1094.

Boberg, M., Croon, L. B., Gustafsson, I. B., & Vessby, B. (1985). Platelet fatty

acid composition in relation to fatty acid composition in plasma and to

serum lipoprotein lipids in healthy subjects with special reference to the

linoleic acid pathway. Clin Sci (Lond), 68(5), 581-587.

Boers, M. (2009). Just released from the ASAS factory! First steps towards a

disease activity score for ankylosing spondylitis. Ann Rheum Dis, 68(1),

1-2.

Bourne, J. T., Kumar, P., Huskisson, E. C., Mageed, R., Unsworth, D. J., &

Wojtulewski, J. A. (1985). Arthritis and coeliac disease. Ann Rheum Dis,

44(9), 592-598.

Boyer, G. S., Lanier, A. P., & Templin, D. W. (1988). Prevalence rates of

spondyloarthropathies, rheumatoid arthritis, and other rheumatic

disorders in an Alaskan Inupiat Eskimo population. J Rheumatol, 15(4),

678-683.

Boyer, G. S., Templin, D. W., Cornoni-Huntley, J. C., Everett, D. F., Lawrence,

R. C., Heyse, S. F. et al. (1994). Prevalence of spondyloarthropathies in

Alaskan Eskimos. J Rheumatol, 21(12), 2292-2297.

40

Braun, J., Brandt, J., Listing, J., Rudwaleit, M., & Sieper, J. (2003). Biologic

therapies in the spondyloarthritis: new opportunities, new challenges.

Curr Opin Rheumatol, 15(4), 394-407.

Bremander, A., Petersson, I. F., Bergman, S., & Englund, M. (2011). Population-

based estimates of common comorbidities and cardiovascular disease in

ankylosing spondylitis. Arthritis Care Res, 63(4), 550-556.

Brunborg, L. A., Madland, T. M., Lind, R. A., Arslan, G., Berstad, A., & Froyland,

L. (2008). Effects of short-term oral administration of dietary marine oils

in patients with inflammatory bowel disease and joint pain: a pilot study

comparing seal oil and cod liver oil. Clin Nutr, 27(4), 614-622.

Burke, B. S. (1947). The dietary history as a tool in research. J Am Diet Assoc,

23(12), 1041-1046.

Calder, P. C. (2006). n-3 polyunsaturated fatty acids, inflammation, and

inflammatory diseases. Am J Clin Nutr, 83(6 Suppl), 1505S-1519S.

Calder, P. C. (2009). Polyunsaturated fatty acids and inflammatory processes:

New twists in an old tale. Biochimie, 91(6), 791-795.

Calin, A., Garrett, S., Whitelock, H., Kennedy, L. G., O’Hea, J., Mallorie, P. et al.

(1994). A new approach to defining functional ability in ankylosing

spondylitis: the development of the Bath Ankylosing Spondylitis

Functional Index. J Rheumatol, 21(12), 2281-2285.

Cartwright, I. J., Pockley, A. G., Galloway, J. H., Greaves, M., & Preston, F. E.

(1985). The effects of dietary omega-3 polyunsaturated fatty acids on

erythrocyte membrane phospholipids, erythrocyte deformability and

blood viscosity in healthy volunteers. Atherosclerosis, 55(3), 267-281.

Chavali, S. R., Zhong, W. W., & Forse, R. A. (1998). Dietary alpha-linolenic acid

increases TNF-alpha, and decreases IL-6, IL-10 in response to LPS:

effects of sesamin on the delta-5 desaturation of omega6 and omega3 fatty

acids in mice. Prostaglandins Leukot Essent Fatty Acids, 58(3), 185-191.

Cleland, L. G., James, M. J., & Proudman, S. M. (2006). Fish oil: what the

prescriber needs to know. Arthritis Res Ther, 8(1), 202.

Cronstedt, H., Waldner, A., & Stenstrom, C. H. (1999). The Swedish version of

the Bath ankylosing spondylitis functional index. Reliability and validity.

Scand J Rheumatol Suppl, 111, 1-9.

41

Cunnane, S. C., & Anderson, M. J. (1997). The majority of dietary linoleate in

growing rats is beta-oxidized or stored in visceral fat. J Nutr, 127(1),

146-152.

De Vos, M., Cuvelier, C., Mielants, H., Veys, E., Barbier, F., & Elewaut, A.

(1989). Ileocolonoscopy in seronegative spondylarthropathy.

Gastroenterology, 96(2 Pt 1), 339-344.

De Vos, M., Mielants, H., Cuvelier, C., Elewaut, A., & Veys, E. (1996). Long-term

evolution of gut inflammation in patients with spondyloarthropathy.

Gastroenterology, 110(6), 1696-1703.

Dik, V. K., Peters, M. J., Dijkmans, P. A., Van der Weijden, M. A., De Vries, M.

K., Dijkmans, B. A. et al. (2010). The relationship between disease-related

characteristics and conduction disturbances in ankylosing spondylitis.

Scand J Rheumatol, 39(1), 38-41.

Ebringer, A., & Wilson, C. (1996). The use of a low starch diet in the treatment

of patients suffering from ankylosing spondylitis. Clin Rheumatol, 15

Suppl 1, 62-66.

Engström-Laurent, A., & Malloy, K. (1994). Reumatologi. Almqvist & Wiksell

Medicin.

Feltelius, N., Hedenstrom, H., Hillerdal, G., & Hallgren, R. (1986). Pulmonary

involvement in ankylosing spondylitis. Ann Rheum Dis, 45(9), 736-740.

Gaal, J., Lakos, G., Szodoray, P., Kiss, J., Horvath, I., Horkay, E. et al. (2009).

Immunological and clinical effects of alphacalcidol in patients with

psoriatic arthropathy: results of an open, follow-up pilot study. Acta

Derm Venereol, 89(2), 140-144.

Galli, C., & Calder, P. C. (2009). Effects of fat and fatty acid intake on

inflammatory and immune responses: a critical review. Ann Nutr Metab,

55(1-3), 123-139.

Garrett, S., Jenkinson, T., Kennedy, L. G., Whitelock, H., Gaisford, P., & Calin,

A. (1994). A new approach to defining disease status in ankylosing

spondylitis: the Bath Ankylosing Spondylitis Disease Activity Index. J

Rheumatol, 21(12), 2286-2291.

Gibson, P. R., & Shepherd, S. J. (2005). Personal view: food for thought--

western lifestyle and susceptibility to Crohn’s disease. The FODMAP

hypothesis. Aliment Pharmacol Ther, 21(12), 1399-1409.

42

Gran, J. T., & Husby, G. (1993). The epidemiology of ankylosing spondylitis.

Semin Arthritis Rheum, 22(5), 319-334.

Gratacos, J., Orellana, C., Sanmarti, R., Sole, M., Collado, A., Gomez-Casanovas,

E. et al. (1997). Secondary amyloidosis in ankylosing spondylitis. A

systematic survey of 137 patients using abdominal fat aspiration. J

Rheumatol, 24(5), 912-915.

Guarner, F. (2007). Studies with inulin-type fructans on intestinal infections,

permeability, and inflammation. J Nutr, 137(11 Suppl), 2568S-2571S.

Hagfors, L. (2003). Mediterranean dietary intervention study of patients with

rheumatoid arthritis. PhD Thesis. Umeå University

Haglund, E., Bremander, A. B., Petersson, I. F., Strombeck, B., Bergman, S.,

Jacobsson, L. T. et al. (2011). Prevalence of spondyloarthritis and its

subtypes in southern Sweden. Ann Rheum Dis, 70(6), 943-948.

Han, C., Robinson, D. W. J., Hackett, M. V., Paramore, L. C., Fraeman, K. H., &

Bala, M. V. (2006). Cardiovascular disease and risk factors in patients

with rheumatoid arthritis, psoriatic arthritis, and ankylosing spondylitis.

J Rheumatol, 33(11), 2167-2172.

Harris, W. S., Pottala, J. V., Sands, S. A., & Jones, P. G. (2007). Comparison of

the effects of fish and fish-oil capsules on the n 3 fatty acid content of

blood cells and plasma phospholipids. Am J Clin Nutr, 86(6), 1621-1625.

Haugen, M., Kjeldsen-Kragh, J., Nordvag, B. Y., & Forre, O. (1991). Diet and

disease symptoms in rheumatic diseases--results of a questionnaire based

survey. Clin Rheumatol, 10(4), 401-407.

Hodson, L., Skeaff, C. M., & Fielding, B. A. (2008). Fatty acid composition of

adipose tissue and blood in humans and its use as a biomarker of dietary

intake. Prog Lipid Res, 47(5), 348-380.

Hollan, I., Saatvedt, K., Almdahl, S. M., Mikkelsen, K., Moer, R., Halvorsen, P.

et al. (2008). Spondyloarthritis: a strong predictor of early coronary

artery bypass grafting. Scand J Rheumatol, 37(1), 18-22.

Horrobin, D. F. (1987). Low prevalences of coronary heart disease (CHD),

psoriasis, asthma and rheumatoid arthritis in Eskimos: are they caused by

high dietary intake of eicosapentaenoic acid (EPA), a genetic variation of

essential fatty acid (EFA) metabolism or a combination of both? Med

Hypotheses, 22(4), 421-428.

43

Hu, F. B., & Willett, W. C. (2002). Optimal diets for prevention of coronary

heart disease. JAMA, 288(20), 2569-2578.

Huckins, D., Felson, D. T., & Holick, M. (1990). Treatment of psoriatic arthritis

with oral 1,25-dihydroxyvitamin D3: a pilot study. Arthritis Rheum,

33(11), 1723-1727.

Inman, R. D., & El-Gabalawy, H. S. (2009). The immunology of ankylosing

spondylitis and rheumatoid arthritis: a tale of similarities and

dissimilarities. Clin Exp Rheumatol, 27(4 Suppl 55), S26-32.

Jenks, K., Stebbings, S., Burton, J., Schultz, M., Herbison, P., & Highton, J.

(2010). Probiotic therapy for the treatment of spondyloarthritis: a

randomized controlled trial. J Rheumatol, 37(10), 2118-2125.

Johansson, G., & Westerterp, K. R. (2008). Assessment of the physical activity

level with two questions: validation with doubly labeled water. Int J Obes

(Lond), 32(6), 1031-1033.

Johansson, I., Hallmans, G., Wikman, A., Biessy, C., Riboli, E., & Kaaks, R.

(2002). Validation and calibration of food-frequency questionnaire

measurements in the Northern Sweden Health and Disease cohort. Public

Health Nutr, 5(3), 487-496.

Jowett, S. L., Seal, C. J., Phillips, E., Gregory, W., Barton, J. R., & Welfare, M. R.

(2004). Dietary beliefs of people with ulcerative colitis and their effect on

relapse and nutrient intake. Clin Nutr, 23(2), 161-170.

Hellgren, J.H. (1963). The epidemiology of chronic rheumatism; a symposium.

Oxford: Blackwell Scientific Publications.

Kallikorm, R., Uibo, O., & Uibo, R. (2000). Coeliac disease in

spondyloarthropathy: usefulness of serological screening. Clin

Rheumatol, 19(2), 118-122.

Kallings, L. V. (2002). Åtgärder mot fetma. Nationell inventering av pågående

studier/projekt.

Karimi, O., Pena, A. S., & van Bodegraven, A. A. (2005). Probiotics (VSL#3) in

arthralgia in patients with ulcerative colitis and Crohn’s disease: a pilot

study. Drugs Today, 41(7), 453-459.

Khan, M. A. (1996). Epidemiology of HLA-B27 and Arthritis. Clin Rheumatol, 15

Suppl 1, 10-12.

44

Kia, K. F., Nair, R. P., Ike, R. W., Hiremagalore, R., Elder, J. T., & Ellis, C. N.

(2007). Prevalence of antigliadin antibodies in patients with psoriasis is

not elevated compared with controls. Am J Clin Dermatol, 8(5), 301-305.

Kiris, A., Ozgocmen, S., Kocakoc, E., Ardicoglu, O., & Ogur, E. (2003). Lung

findings on high resolution CT in early ankylosing spondylitis. Eur J

Radiol, 47(1), 71-76.

Kromann, N., & Green, A. (1980). Epidemiological studies in the Upernavik

district, Greenland. Incidence of some chronic diseases 1950-1974. Acta

Med Scand, 208(5), 401-406.

Lairon, D. (2007). Intervention studies on Mediterranean diet and

cardiovascular risk. Mol Nutr Food Res, 51(10), 1209-1214.

Lange, U., Jung, O., Teichmann, J., & Neeck, G. (2001). Relationship between

disease activity and serum levels of vitamin D metabolites and

parathyroid hormone in ankylosing spondylitis. Osteoporos Int, 12(12),

1031-1035.

Lange, U., Teichmann, J., Strunk, J., Muller-Ladner, U., & Schmidt, K. L.

(2005). Association of 1.25 vitamin D3 deficiency, disease activity and low

bone mass in ankylosing spondylitis. Osteoporos Int, 16(12), 1999-2004.

Lassus, A., Dahlgren, A. L., Halpern, M. J., Santalahti, J., & Happonen, H. P.

(1990). Effects of dietary supplementation with polyunsaturated ethyl

ester lipids (Angiosan) in patients with psoriasis and psoriatic arthritis. J

Int Med Res, 18(1), 68-73.

Lawrence, R. C., Everett, D. F., Benevolenskaya, L. I., Boyer, G. S., Erdesz, S.,

Templin, D. W. et al. (1996). Spondyloarthropathies in circumpolar

populations: I. Design and methods of United States and Russian studies.

Arctic Med Res, 55(4), 187-194.

Lindqvist, U., Rudsander, A., Bostrom, A., Nilsson, B., & Michaelsson, G.

(2002). IgA antibodies to gliadin and coeliac disease in psoriatic arthritis.

Rheumatology (Oxford), 41(1), 31-37.

Lories, R. J., & Baeten, D. L. (2009). Differences in pathophysiology between

rheumatoid arthritis and ankylosing spondylitis. Clin Exp Rheumatol,

27(4 Suppl 55), S10-4.

45

Lukas, C., Landewe, R., Sieper, J., Dougados, M., Davis, J., Braun, J. et al.

(2009). Development of an ASAS-endorsed disease activity score

(ASDAS) in patients with ankylosing spondylitis. Ann Rheum Dis, 68(1),

18-24.

Machado, P. M., Landewe, R. B., & van der Heijde, D. M. (2011). Endorsement

of Definitions of Disease Activity States and Improvement Scores for the

Ankylosing Spondylitis Disease Activity Score: Results from OMERACT

10. J Rheumatol, 38(7), 1502-1506.

Madland, T. M., Bjorkkjaer, T., Brunborg, L. A., Froyland, L., Berstad, A., &

Brun, J. G. (2006). Subjective improvement in patients with psoriatic

arthritis after short-term oral treatment with seal oil. A pilot study with

double blind comparison to soy oil. J Rheumatol, 33(2), 307-310.

Maksymowych, W. P., Landewe, R., Conner-Spady, B., Dougados, M., Mielants,

H., van der Tempel, H. et al. (2007). Serum matrix metalloproteinase 3 is

an independent predictor of structural damage progression in patients

with ankylosing spondylitis. Arthritis Rheum, 56(6), 1846-1853.

Malcom, G. T., Bhattacharyya, A. K., Velez-Duran, M., Guzman, M. A.,

Oalmann, M. C., & Strong, J. P. (1989). Fatty acid composition of adipose

tissue in humans: differences between subcutaneous sites. Am J Clin

Nutr, 50(2), 288-291.

Martinelli, N., Girelli, D., Malerba, G., Guarini, P., Illig, T., Trabetti, E. et al.

(2008). FADS genotypes and desaturase activity estimated by the ratio of

arachidonic acid to linoleic acid are associated with inflammation and

coronary artery disease. Am J Clin Nutr, 88(4), 941-949.

Masi AT, Aldag JC, Mohan PC. (1999). Determinants of significantly lower

serum total cholesterol levels in ankylosing spondylitis patients than age-,

gender-, and medical service matched control patients: results of

multivariate analysis. Arthritis Rheum, 42 (Suppl.):S300

Masi AT, Aldag JC, Mohan PC. (2000). Significantly lower serum triglyceride

levels in ankylosing spondylitis patients than age-, gender-, and medical

service matched controls: results of multivariate analysis. Arthritis

Rheum, 43 (Suppl.):S102

46

McGettigan, P., & Henry, D. (2006). Cardiovascular risk and inhibition of

cyclooxygenase: a systematic review of the observational studies of

selective and nonselective inhibitors of cyclooxygenase 2. JAMA, 296(13),

1633-1644.

Merino, D. M., Ma, D. W., & Mutch, D. M. (2010). Genetic variation in lipid

desaturases and its impact on the development of human disease. Lipids

Health Dis, 9, 63.

Mermerci Baskan, B., Pekin Dogan, Y., Sivas, F., Bodur, H., & Ozoran, K.

(2010). The relation between osteoporosis and vitamin D levels and

disease activity in ankylosing spondylitis. Rheumatol Int, 30(3), 375-381.

Mielants, H., Veys, E. M., Cuvelier, C., & De Vos, M. (1996). Course of gut

inflammation in spondylarthropathies and therapeutic consequences.

Baillieres Clin Rheumatol, 10(1), 147-164.

Mielants, H., Veys, E. M., Goemaere, S., Goethals, K., Cuvelier, C., & De Vos, M.

(1991). Gut inflammation in the spondyloarthropathies: clinical,

radiologic, biologic and genetic features in relation to the type of

histology. A prospective study. J Rheumatol, 18(10), 1542-1551.

Momeni, M., Taylor, N., & Tehrani, M. (2011). Cardiopulmonary manifestations

of ankylosing spondylitis. Int J Rheumatol, 2011, 728471.

Mullaji, A. B., Upadhyay, S. S., & Ho, E. K. (1994). Bone mineral density in

ankylosing spondylitis. DEXA comparison of control subjects with mild

and advanced cases. J Bone Joint Surg Br, 76(4), 660-665.

Obukowicz, M. G., Raz, A., Pyla, P. D., Rico, J. G., Wendling, J. M., &

Needleman, P. (1998). Identification and characterization of a novel

delta6/delta5 fatty acid desaturase inhibitor as a potential anti-

inflammatory agent. Biochem Pharmacol, 55(7), 1045-1058.

Peterson, K., McDonagh, M., Thakurta, S., & Dana., T. (2010). Drug Class

Review: Nonsteroidal Antiinflammatory Drugs (NSAIDs): Final Update 4

Report. ncbi.nlm.nih.gov.

Psota, T. L., Gebauer, S. K., & Kris-Etherton, P. (2006). Dietary omega-3 fatty

acid intake and cardiovascular risk. The American journal of cardiology,

98(4), 3-18.

47

Ratnayake, W. M., & Galli, C. (2009). Fat and fatty acid terminology, methods of

analysis and fat digestion and metabolism: a background review paper.

Ann Nutr Metab, 55(1-3), 8-43.

Richardson, E. S., Iaizzo, P. A., & Xiao, Y. F. (2011). Electrophysiological

mechanisms of the anti-arrhythmic effects of omega-3 fatty acids. J

Cardiovasc Transl Res, 4(1), 42-52.

Riente, L., Chimenti, D., Pratesi, F., Delle Sedie, A., Tommasi, S., Tommasi, C. et

al. (2004). Antibodies to tissue transglutaminase and Saccharomyces

cerevisiae in ankylosing spondylitis and psoriatic arthritis. J Rheumatol,

31(5), 920-924.

Roldan, C. A., Chavez, J., Wiest, P. W., Qualls, C. R., & Crawford, M. H. (1998).

Aortic root disease and valve disease associated with ankylosing

spondylitis. J Am Coll Cardiol, 32(5), 1397-1404.

Rose, H. M., Ragan, C., & et, a. (1948). Differential agglutination of normal and

sensitized sheep erythrocytes by sera of patients with rheumatoid

arthritis. Proc Soc Exp Biol Med, 68(1), 1-6.

Rosenblum, H., & Amital, H. (2011). Anti-TNF therapy: Safety aspects of taking

the risk. Autoimmun Rev, 10(9), 563-568.

Roubenoff, R., Roubenoff, R. A., Cannon, J. G., Kehayias, J. J., Zhuang, H.,

Dawson-Hughes, B. et al. (1994). Rheumatoid cachexia: cytokine-driven

hypermetabolism accompanying reduced body cell mass in chronic

inflammation. J Clin Invest, 93(6), 2379-2386.

Rudwaleit, M., & Baeten, D. (2006). Ankylosing spondylitis and bowel disease.

Best Pract Res Clin Rheumatol, 20(3), 451-471.

Rudwaleit, M., & Taylor, W. J. (2010). Classification criteria for psoriatic

arthritis and ankylosing spondylitis/axial spondyloarthritis. Best Pract

Res Clin Rheumatol, 24(5), 589-604.

Sanders, T. A. (2009). Fat and fatty acid intake and metabolic effects in the

human body. Ann Nutr Metab, 55(1-3), 162-172.

Senocak, O., Manisali, M., Ozaksoy, D., Sevinc, C., & Akalin, E. (2003). Lung

parenchyma changes in ankylosing spondylitis: demonstration with high

resolution CT and correlation with disease duration. Eur J Radiol, 45(2),

117-122.

48

Sieper, J., Rudwaleit, M., Khan, MA., Braun, J. (2006). Concepts and

epidemiology of spondyloarthritis. Best Pract Res Clin Rheumatol,

20:401-417.

Sheehan, N. J., Slavin, B. M., Donovan, M. P., Mount, J. N., & Mathews, J. A.

(1986). Lack of correlation between clinical disease activity and

erythrocyte sedimentation rate, acute phase proteins or protease

inhibitors in ankylosing spondylitis. Br J Rheumatol, 25(2), 171-174.

Shimizu, S., Akimoto, K., Shinmen, Y., Kawashima, H., Sugano, M., & Yamada,

H. (1991). Sesamin is a potent and specific inhibitor of delta 5 desaturase

in polyunsaturated fatty acid biosynthesis. Lipids, 26(7), 512-516.

Shinebaum, R., Neumann, V., Hopkins, R., Cooke, E. M., & Wright, V. (1984).

Attempt to modify klebsiella carriage in ankylosing spondylitic patients by

diet: correlation of klebsiella carriage with disease activity. Ann Rheum

Dis, 43(2), 196-199.

Sinclair, H. M. (1953). The diet of canadian indians and eskimos. Proceedings of

the Nutrition Society, 12(01), 69-82.

Singh, G., Kumari, N., Aggarwal, A., Krishnani, N., & Misra, R. (2007).

Prevalence of subclinical amyloidosis in ankylosing spondylitis. J

Rheumatol, 34(2), 371-373.

Spoorenberg, A., van der Heijde, D., de Klerk, E., Dougados, M., de Vlam, K.,

Mielants, H. et al. (1999). Relative value of erythrocyte sedimentation rate

and C-reactive protein in assessment of disease activity in ankylosing

spondylitis. J Rheumatol, 26(4), 980-984.

Staal-van den Brekel, A. J., Dentener, M. A., Schols, A. M., Buurman, W. A., &

Wouters, E. F. (1995). Increased resting energy expenditure and weight

loss are related to a systemic inflammatory response in lung cancer

patients. J Clin Oncol, 13(10), 2600-2605.

Stebbings S. (2009). The severity of bowel symptoms is similar in active

ankylosing spondylitis and crohn’s disease. Ann Rheum Dis, 68 (Suppl3):

649.

Stegmayr, B., Lundberg, V., & Asplund, K. (2003). The events registration and

survey procedures in the Northern Sweden MONICA Project. Scand J

Public Health Suppl, 61, 9-17.

49

Teichmann, J., Voglau, M. J., & Lange, U. (2009). Antibodies to human tissue

transglutaminase and alterations of vitamin D metabolism in ankylosing

spondylitis and psoriatic arthritis. Rheumatol Int, 30(12):1559-63

Ten Bruggencate, S. J., Bovee-Oudenhoven, I. M., Lettink-Wissink, M. L.,

Katan, M. B., & Van Der Meer, R. (2004). Dietary fructo-oligosaccharides

and inulin decrease resistance of rats to salmonella: protective role of

calcium. Gut, 53(4), 530-535.

Togrol, R. E., Nalbant, S., Solmazgul, E., Ozyurt, M., Kaplan, M., Kiralp, M. Z. et

al. (2009). The Significance of Coeliac Disease Antibodies in Patients with

Ankylosing Spondylitis: a Case-controlled Study. J Int Med Res, 37(1),

220-226.

Torres, M. I., & Rios, A. (2008). Current view of the immunopathogenesis in

inflammatory bowel disease and its implications for therapy. World J

Gastroenterol, 14(13), 1972-1980.

Utsunomiya, T., Chavali, S. R., Zhong, W. W., & Forse, R. A. (2000). Effects of

sesamin-supplemented dietary fat emulsions on the ex vivo production of

lipopolysaccharide-induced prostanoids and tumor necrosis factor alpha

in rats. Am J Clin Nutr, 72(3), 804-808.

van Halm, V. P., van Denderen, J. C., Peters, M. J., Twisk, J. W., van der Paardt,

M., van der Horst-Bruinsma, I. E. et al. (2006). Increased disease activity

is associated with a deteriorated lipid profile in patients with ankylosing

spondylitis. Ann Rheum Dis, 65(11), 1473-1477.

van der Heijde, D., Lie, E., Kvien, T. K., Sieper, J., Van den Bosch, F., Listing, J.

et al. (2008). The ASDAS is a highly discriminatory ASAS-endorsed

disease activity score in patients with ankylosing spondylitis. Ann Rheum

Dis, 68(12):1811-8

van der Linden, S. (2008). Issues in the treatment of ankylosing spondylitis

with non-steroidal anti-inflammatory drugs. Wien Med Wochenschr,

158(7-8), 195-199.

van der Linden, S., Valkenburg, H. A., & Cats, A. (1984). Evaluation of

diagnostic criteria for ankylosing spondylitis. A proposal for modification

of the New York criteria. Arthritis Rheum, 27(4), 361-368.

50

Vonk, R. J., Priebe, M. G., Koetse, H. A., Stellaard, F., Lenoir-Wijnkoop, I., Antoine, J. M. et al. (2003). Lactose intolerance: analysis of underlying factors. Eur J Clin Invest, 33(1), 70-75.

Waldner, A., Cronstedt, H., & Stenstrom, C. H. (1999). The Swedish version of the Bath ankylosing spondylitis disease activity index. Reliability and validity. Scand J Rheumatol Suppl, 111, 10-16.

Waterhouse, J. C., Perez, T. H., & Albert, P. J. (2009). Reversing bacteria-induced vitamin D receptor dysfunction is key to autoimmune disease. Ann N Y Acad Sci, 1173, 757-765.

Weinehall, L., Hallgren, C. G., Westman, G., Janlert, U., & Wall, S. (1998). Reduction of selection bias in primary prevention of cardiovascular disease through involvement of primary health care. Scand J Prim Health Care, 16(3), 171-176.

Weinehall, L., Hellsten, G., Boman, K., & Hallmans, G. (2001). Prevention of cardiovascular disease in Sweden: the Norsjo community intervention programme--motives, methods and intervention components. Scand J Public Health Suppl, 56, 13-20.

Wendling, D., Cedoz, J. P., & Racadot, E. (2008). Serum levels of MMP-3 and cathepsin K in patients with ankylosing spondylitis: effect of TNFalpha antagonist therapy. Joint Bone Spine, 75(5), 559-562.

Wennberg, M., Vessby, B., & Johansson, I. (2009). Evaluation of relative intake of fatty acids according to the Northern Sweden FFQ with fatty acid levels in erythrocyte membranes as biomarkers. Public Health Nutr, 12(9), 1477-1484.

Wiehl, D. G. (1942). Diets of a group of aircraft workers in Southern California. The Milbank Memorial Fund Quarterly, 329-366.

Will, R., Palmer, R., Bhalla, A. K., Ring, F., & Calin, A. (1989). Osteoporosis in early ankylosing spondylitis: a primary pathological event? Lancet, 2(8678-8679), 1483-1485.

Yu. (2000). Spondyloarthropathies. Kelley’s textbook of rheumatology. WB Saunders Company, Philadelphia.

Zochling, J., & Braun, J. (2009). Mortality in rheumatoid arthritis and ankylosing spondylitis. Clin Exp Rheumatol, 27(4 Suppl 55), S127-30.

Zochling, J., van der Heijde, D., Burgos-Vargas, R., Collantes, E., Davis, J. C. J., Dijkmans, B. et al. (2006). ASAS/EULAR recommendations for the management of ankylosing spondylitis. Ann Rheum Dis, 65(4), 442-452.

Department of Public health and clinical medicineDivision of Rheumatology Umeå University, 901 87 Umeå, Swedenwww.umu.se

ISSN 0346-6612ISBN 978-91-7459-272-6